{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.5,
  "eval_steps": 500,
  "global_step": 287,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "completion_length": 243.21429443359375,
      "epoch": 0.0017421602787456446,
      "grad_norm": 11.870542526245117,
      "kl": 0.0,
      "learning_rate": 1e-06,
      "loss": -0.0,
      "reward": 1.1250000596046448,
      "reward_std": 0.35204461216926575,
      "rewards/accuracy_reward": 0.2321428693830967,
      "rewards/format_reward": 0.892857164144516,
      "step": 1
    },
    {
      "completion_length": 232.3928680419922,
      "epoch": 0.003484320557491289,
      "grad_norm": 21.013626098632812,
      "kl": 0.002597808837890625,
      "learning_rate": 1e-06,
      "loss": 0.0001,
      "reward": 1.0535714626312256,
      "reward_std": 0.4149572253227234,
      "rewards/accuracy_reward": 0.196428582072258,
      "rewards/format_reward": 0.8571429252624512,
      "step": 2
    },
    {
      "completion_length": 314.5893020629883,
      "epoch": 0.005226480836236934,
      "grad_norm": 8.975765228271484,
      "kl": 0.01239013671875,
      "learning_rate": 1e-06,
      "loss": 0.0005,
      "reward": 1.0178571939468384,
      "reward_std": 0.2851574905216694,
      "rewards/accuracy_reward": 0.1071428656578064,
      "rewards/format_reward": 0.910714328289032,
      "step": 3
    },
    {
      "completion_length": 284.1964416503906,
      "epoch": 0.006968641114982578,
      "grad_norm": 4.65426778793335,
      "kl": 0.008523941040039062,
      "learning_rate": 1e-06,
      "loss": 0.0003,
      "reward": 1.160714328289032,
      "reward_std": 0.2610500454902649,
      "rewards/accuracy_reward": 0.196428582072258,
      "rewards/format_reward": 0.9642857313156128,
      "step": 4
    },
    {
      "completion_length": 298.21429443359375,
      "epoch": 0.008710801393728223,
      "grad_norm": 1.674553632736206,
      "kl": 0.00372314453125,
      "learning_rate": 1e-06,
      "loss": 0.0001,
      "reward": 1.1250000596046448,
      "reward_std": 0.3193712383508682,
      "rewards/accuracy_reward": 0.1785714402794838,
      "rewards/format_reward": 0.9464285969734192,
      "step": 5
    },
    {
      "completion_length": 286.9464340209961,
      "epoch": 0.010452961672473868,
      "grad_norm": 1.6619535684585571,
      "kl": 0.000659942626953125,
      "learning_rate": 1e-06,
      "loss": 0.0,
      "reward": 1.2678571939468384,
      "reward_std": 0.3606105446815491,
      "rewards/accuracy_reward": 0.2857142984867096,
      "rewards/format_reward": 0.9821428656578064,
      "step": 6
    },
    {
      "completion_length": 323.1964416503906,
      "epoch": 0.012195121951219513,
      "grad_norm": 1.0895068645477295,
      "kl": 0.005126953125,
      "learning_rate": 1e-06,
      "loss": 0.0002,
      "reward": 1.1071429252624512,
      "reward_std": 0.1428571529686451,
      "rewards/accuracy_reward": 0.1428571492433548,
      "rewards/format_reward": 0.9642857313156128,
      "step": 7
    },
    {
      "completion_length": 288.1428756713867,
      "epoch": 0.013937282229965157,
      "grad_norm": 1.8034625053405762,
      "kl": 0.0026731491088867188,
      "learning_rate": 1e-06,
      "loss": 0.0001,
      "reward": 1.2142857909202576,
      "reward_std": 0.2967643216252327,
      "rewards/accuracy_reward": 0.232142873108387,
      "rewards/format_reward": 0.9821428656578064,
      "step": 8
    },
    {
      "completion_length": 263.23216247558594,
      "epoch": 0.0156794425087108,
      "grad_norm": 0.7990542054176331,
      "kl": 0.005615234375,
      "learning_rate": 1e-06,
      "loss": 0.0002,
      "reward": 1.1071429252624512,
      "reward_std": 0.18409645557403564,
      "rewards/accuracy_reward": 0.1428571492433548,
      "rewards/format_reward": 0.9642857313156128,
      "step": 9
    },
    {
      "completion_length": 201.53572845458984,
      "epoch": 0.017421602787456445,
      "grad_norm": 1.5415644645690918,
      "kl": 0.0062713623046875,
      "learning_rate": 1e-06,
      "loss": 0.0003,
      "reward": 1.3571429252624512,
      "reward_std": 0.2363857999444008,
      "rewards/accuracy_reward": 0.3571428656578064,
      "rewards/format_reward": 1.0,
      "step": 10
    },
    {
      "completion_length": 73.30357551574707,
      "epoch": 0.01916376306620209,
      "grad_norm": 1.2912265062332153,
      "kl": 0.011810302734375,
      "learning_rate": 1e-06,
      "loss": 0.0005,
      "reward": 1.2500000596046448,
      "reward_std": 0.1428571529686451,
      "rewards/accuracy_reward": 0.25000000931322575,
      "rewards/format_reward": 1.0,
      "step": 11
    },
    {
      "completion_length": 131.8214340209961,
      "epoch": 0.020905923344947737,
      "grad_norm": 0.8542214035987854,
      "kl": 0.009796142578125,
      "learning_rate": 1e-06,
      "loss": 0.0004,
      "reward": 1.1250000596046448,
      "reward_std": 0.13981622830033302,
      "rewards/accuracy_reward": 0.160714291036129,
      "rewards/format_reward": 0.9642857313156128,
      "step": 12
    },
    {
      "completion_length": 133.25000762939453,
      "epoch": 0.02264808362369338,
      "grad_norm": 1.5221061706542969,
      "kl": 0.014312744140625,
      "learning_rate": 1e-06,
      "loss": 0.0006,
      "reward": 1.2142857313156128,
      "reward_std": 0.2253357619047165,
      "rewards/accuracy_reward": 0.2142857313156128,
      "rewards/format_reward": 1.0,
      "step": 13
    },
    {
      "completion_length": 83.32143020629883,
      "epoch": 0.024390243902439025,
      "grad_norm": 0.7153174877166748,
      "kl": 0.0189208984375,
      "learning_rate": 1e-06,
      "loss": 0.0008,
      "reward": 1.1964285969734192,
      "reward_std": 0.07695358991622925,
      "rewards/accuracy_reward": 0.1964285746216774,
      "rewards/format_reward": 1.0,
      "step": 14
    },
    {
      "completion_length": 96.44642925262451,
      "epoch": 0.02613240418118467,
      "grad_norm": 1.98847496509552,
      "kl": 0.01861572265625,
      "learning_rate": 1e-06,
      "loss": 0.0007,
      "reward": 1.071428656578064,
      "reward_std": 0.1428571492433548,
      "rewards/accuracy_reward": 0.0892857201397419,
      "rewards/format_reward": 0.9821428656578064,
      "step": 15
    },
    {
      "completion_length": 100.4285774230957,
      "epoch": 0.027874564459930314,
      "grad_norm": 1.2997009754180908,
      "kl": 0.0174560546875,
      "learning_rate": 1e-06,
      "loss": 0.0007,
      "reward": 1.2321429252624512,
      "reward_std": 0.14838216826319695,
      "rewards/accuracy_reward": 0.23214287497103214,
      "rewards/format_reward": 1.0,
      "step": 16
    },
    {
      "completion_length": 126.35714721679688,
      "epoch": 0.029616724738675958,
      "grad_norm": 1.5723376274108887,
      "kl": 0.01885986328125,
      "learning_rate": 1e-06,
      "loss": 0.0008,
      "reward": 1.3214285969734192,
      "reward_std": 0.24695908278226852,
      "rewards/accuracy_reward": 0.3392857313156128,
      "rewards/format_reward": 0.9821428656578064,
      "step": 17
    },
    {
      "completion_length": 128.23215103149414,
      "epoch": 0.0313588850174216,
      "grad_norm": 1.6021169424057007,
      "kl": 0.016815185546875,
      "learning_rate": 1e-06,
      "loss": 0.0007,
      "reward": 1.2142857909202576,
      "reward_std": 0.18409645557403564,
      "rewards/accuracy_reward": 0.2321428656578064,
      "rewards/format_reward": 0.9821428656578064,
      "step": 18
    },
    {
      "completion_length": 103.39286041259766,
      "epoch": 0.033101045296167246,
      "grad_norm": 1.91715407371521,
      "kl": 0.028564453125,
      "learning_rate": 1e-06,
      "loss": 0.0011,
      "reward": 1.1250000596046448,
      "reward_std": 0.21981074661016464,
      "rewards/accuracy_reward": 0.1428571492433548,
      "rewards/format_reward": 0.9821428656578064,
      "step": 19
    },
    {
      "completion_length": 91.7500057220459,
      "epoch": 0.03484320557491289,
      "grad_norm": 2.3495121002197266,
      "kl": 0.0179443359375,
      "learning_rate": 1e-06,
      "loss": 0.0007,
      "reward": 1.1071428656578064,
      "reward_std": 0.11266788095235825,
      "rewards/accuracy_reward": 0.12500000931322575,
      "rewards/format_reward": 0.9821428656578064,
      "step": 20
    },
    {
      "completion_length": 85.98214721679688,
      "epoch": 0.036585365853658534,
      "grad_norm": 2.7929649353027344,
      "kl": 0.0169677734375,
      "learning_rate": 1e-06,
      "loss": 0.0007,
      "reward": 1.2142857313156128,
      "reward_std": 0.33800363540649414,
      "rewards/accuracy_reward": 0.2321428656578064,
      "rewards/format_reward": 0.9821428656578064,
      "step": 21
    },
    {
      "completion_length": 108.64286422729492,
      "epoch": 0.03832752613240418,
      "grad_norm": 1.8589195013046265,
      "kl": 0.016265869140625,
      "learning_rate": 1e-06,
      "loss": 0.0007,
      "reward": 1.1964285969734192,
      "reward_std": 0.2937234044075012,
      "rewards/accuracy_reward": 0.2142857164144516,
      "rewards/format_reward": 0.9821428656578064,
      "step": 22
    },
    {
      "completion_length": 110.83928680419922,
      "epoch": 0.04006968641114982,
      "grad_norm": 2.335958957672119,
      "kl": 0.01531982421875,
      "learning_rate": 1e-06,
      "loss": 0.0006,
      "reward": 1.1964285969734192,
      "reward_std": 0.23086076974868774,
      "rewards/accuracy_reward": 0.2142857164144516,
      "rewards/format_reward": 0.9821428656578064,
      "step": 23
    },
    {
      "completion_length": 137.08929061889648,
      "epoch": 0.041811846689895474,
      "grad_norm": 1.5521477460861206,
      "kl": 0.02178955078125,
      "learning_rate": 1e-06,
      "loss": 0.0009,
      "reward": 1.0892857909202576,
      "reward_std": 0.17651408910751343,
      "rewards/accuracy_reward": 0.1071428619325161,
      "rewards/format_reward": 0.9821428656578064,
      "step": 24
    },
    {
      "completion_length": 79.53571891784668,
      "epoch": 0.04355400696864112,
      "grad_norm": 1.3790969848632812,
      "kl": 0.0279541015625,
      "learning_rate": 1e-06,
      "loss": 0.0011,
      "reward": 1.2142857909202576,
      "reward_std": 0.19514649361371994,
      "rewards/accuracy_reward": 0.2142857238650322,
      "rewards/format_reward": 1.0,
      "step": 25
    },
    {
      "completion_length": 218.87500762939453,
      "epoch": 0.04529616724738676,
      "grad_norm": 0.9893456101417542,
      "kl": 0.014007568359375,
      "learning_rate": 1e-06,
      "loss": 0.0006,
      "reward": 1.160714328289032,
      "reward_std": 0.1785714402794838,
      "rewards/accuracy_reward": 0.196428582072258,
      "rewards/format_reward": 0.9642857313156128,
      "step": 26
    },
    {
      "completion_length": 65.67857551574707,
      "epoch": 0.047038327526132406,
      "grad_norm": 1.9509761333465576,
      "kl": 0.02667236328125,
      "learning_rate": 1e-06,
      "loss": 0.0011,
      "reward": 1.1428571939468384,
      "reward_std": 0.18409645557403564,
      "rewards/accuracy_reward": 0.1428571492433548,
      "rewards/format_reward": 1.0,
      "step": 27
    },
    {
      "completion_length": 103.58928680419922,
      "epoch": 0.04878048780487805,
      "grad_norm": 1.6537208557128906,
      "kl": 0.03692626953125,
      "learning_rate": 1e-06,
      "loss": 0.0015,
      "reward": 1.2500000596046448,
      "reward_std": 0.2253357544541359,
      "rewards/accuracy_reward": 0.2857142984867096,
      "rewards/format_reward": 0.9642857313156128,
      "step": 28
    },
    {
      "completion_length": 48.07143211364746,
      "epoch": 0.050522648083623695,
      "grad_norm": 2.2078816890716553,
      "kl": 0.03143310546875,
      "learning_rate": 1e-06,
      "loss": 0.0013,
      "reward": 1.321428656578064,
      "reward_std": 0.2253357619047165,
      "rewards/accuracy_reward": 0.3392857313156128,
      "rewards/format_reward": 0.9821428656578064,
      "step": 29
    },
    {
      "completion_length": 55.160715103149414,
      "epoch": 0.05226480836236934,
      "grad_norm": 2.4559199810028076,
      "kl": 0.03167724609375,
      "learning_rate": 1e-06,
      "loss": 0.0013,
      "reward": 1.2321429252624512,
      "reward_std": 0.2610500305891037,
      "rewards/accuracy_reward": 0.2321428656578064,
      "rewards/format_reward": 1.0,
      "step": 30
    },
    {
      "completion_length": 131.6964340209961,
      "epoch": 0.05400696864111498,
      "grad_norm": 1.1578391790390015,
      "kl": 0.02935791015625,
      "learning_rate": 1e-06,
      "loss": 0.0012,
      "reward": 1.160714328289032,
      "reward_std": 0.20670335739850998,
      "rewards/accuracy_reward": 0.1964285857975483,
      "rewards/format_reward": 0.9642857313156128,
      "step": 31
    },
    {
      "completion_length": 46.571431159973145,
      "epoch": 0.05574912891986063,
      "grad_norm": 3.844055652618408,
      "kl": 0.0350341796875,
      "learning_rate": 1e-06,
      "loss": 0.0014,
      "reward": 1.3214285969734192,
      "reward_std": 0.19514648616313934,
      "rewards/accuracy_reward": 0.321428582072258,
      "rewards/format_reward": 1.0,
      "step": 32
    },
    {
      "completion_length": 66.50000190734863,
      "epoch": 0.05749128919860627,
      "grad_norm": 1.843333125114441,
      "kl": 0.02716064453125,
      "learning_rate": 1e-06,
      "loss": 0.0011,
      "reward": 1.1964285969734192,
      "reward_std": 0.14838216826319695,
      "rewards/accuracy_reward": 0.196428582072258,
      "rewards/format_reward": 1.0,
      "step": 33
    },
    {
      "completion_length": 123.50000762939453,
      "epoch": 0.059233449477351915,
      "grad_norm": 1.1580967903137207,
      "kl": 0.027099609375,
      "learning_rate": 1e-06,
      "loss": 0.0011,
      "reward": 1.160714328289032,
      "reward_std": 0.1071428619325161,
      "rewards/accuracy_reward": 0.17857144214212894,
      "rewards/format_reward": 0.9821428656578064,
      "step": 34
    },
    {
      "completion_length": 145.44643783569336,
      "epoch": 0.06097560975609756,
      "grad_norm": 0.6297839879989624,
      "kl": 0.02728271484375,
      "learning_rate": 1e-06,
      "loss": 0.0011,
      "reward": 1.1250000596046448,
      "reward_std": 0.1071428619325161,
      "rewards/accuracy_reward": 0.14285714365541935,
      "rewards/format_reward": 0.9821428656578064,
      "step": 35
    },
    {
      "completion_length": 175.9285774230957,
      "epoch": 0.0627177700348432,
      "grad_norm": 1.174981951713562,
      "kl": 0.023681640625,
      "learning_rate": 1e-06,
      "loss": 0.0009,
      "reward": 1.2500000596046448,
      "reward_std": 0.2580091208219528,
      "rewards/accuracy_reward": 0.3035714328289032,
      "rewards/format_reward": 0.9464285969734192,
      "step": 36
    },
    {
      "completion_length": 122.89286041259766,
      "epoch": 0.06445993031358885,
      "grad_norm": 0.48167216777801514,
      "kl": 0.03033447265625,
      "learning_rate": 1e-06,
      "loss": 0.0012,
      "reward": 1.3750000596046448,
      "reward_std": 0.13981622830033302,
      "rewards/accuracy_reward": 0.4107142984867096,
      "rewards/format_reward": 0.9642857313156128,
      "step": 37
    },
    {
      "completion_length": 54.08928871154785,
      "epoch": 0.06620209059233449,
      "grad_norm": 2.2061703205108643,
      "kl": 0.02490234375,
      "learning_rate": 1e-06,
      "loss": 0.001,
      "reward": 1.2142857313156128,
      "reward_std": 0.11266787722706795,
      "rewards/accuracy_reward": 0.2142857164144516,
      "rewards/format_reward": 1.0,
      "step": 38
    },
    {
      "completion_length": 90.75000381469727,
      "epoch": 0.06794425087108014,
      "grad_norm": 1.4908874034881592,
      "kl": 0.02630615234375,
      "learning_rate": 1e-06,
      "loss": 0.0011,
      "reward": 1.160714328289032,
      "reward_std": 0.1181928962469101,
      "rewards/accuracy_reward": 0.1607142984867096,
      "rewards/format_reward": 1.0,
      "step": 39
    },
    {
      "completion_length": 76.19643211364746,
      "epoch": 0.06968641114982578,
      "grad_norm": 1.1106164455413818,
      "kl": 0.03558349609375,
      "learning_rate": 1e-06,
      "loss": 0.0014,
      "reward": 1.1428571939468384,
      "reward_std": 0.1428571529686451,
      "rewards/accuracy_reward": 0.160714291036129,
      "rewards/format_reward": 0.9821428656578064,
      "step": 40
    },
    {
      "completion_length": 93.75000381469727,
      "epoch": 0.07142857142857142,
      "grad_norm": 2.4737675189971924,
      "kl": 0.02557373046875,
      "learning_rate": 1e-06,
      "loss": 0.001,
      "reward": 1.2142857909202576,
      "reward_std": 0.21676981449127197,
      "rewards/accuracy_reward": 0.232142873108387,
      "rewards/format_reward": 0.9821428656578064,
      "step": 41
    },
    {
      "completion_length": 87.07143592834473,
      "epoch": 0.07317073170731707,
      "grad_norm": 2.791724443435669,
      "kl": 0.0623779296875,
      "learning_rate": 1e-06,
      "loss": 0.0025,
      "reward": 1.410714328289032,
      "reward_std": 0.1785714402794838,
      "rewards/accuracy_reward": 0.4107142984867096,
      "rewards/format_reward": 1.0,
      "step": 42
    },
    {
      "completion_length": 71.01786041259766,
      "epoch": 0.07491289198606271,
      "grad_norm": 3.068845748901367,
      "kl": 0.02935791015625,
      "learning_rate": 1e-06,
      "loss": 0.0012,
      "reward": 1.1428571939468384,
      "reward_std": 0.1539071835577488,
      "rewards/accuracy_reward": 0.16071429289877415,
      "rewards/format_reward": 0.9821428656578064,
      "step": 43
    },
    {
      "completion_length": 102.42857360839844,
      "epoch": 0.07665505226480836,
      "grad_norm": 3.0162665843963623,
      "kl": 0.026123046875,
      "learning_rate": 1e-06,
      "loss": 0.001,
      "reward": 1.2500000596046448,
      "reward_std": 0.2142857313156128,
      "rewards/accuracy_reward": 0.2500000149011612,
      "rewards/format_reward": 1.0,
      "step": 44
    },
    {
      "completion_length": 140.58929443359375,
      "epoch": 0.078397212543554,
      "grad_norm": 3.6954751014709473,
      "kl": 0.032470703125,
      "learning_rate": 1e-06,
      "loss": 0.0013,
      "reward": 1.2321429252624512,
      "reward_std": 0.1785714402794838,
      "rewards/accuracy_reward": 0.2857142984867096,
      "rewards/format_reward": 0.9464285969734192,
      "step": 45
    },
    {
      "completion_length": 105.83929061889648,
      "epoch": 0.08013937282229965,
      "grad_norm": 1.9819762706756592,
      "kl": 0.04046630859375,
      "learning_rate": 1e-06,
      "loss": 0.0016,
      "reward": 1.3928571939468384,
      "reward_std": 0.19514650478959084,
      "rewards/accuracy_reward": 0.392857164144516,
      "rewards/format_reward": 1.0,
      "step": 46
    },
    {
      "completion_length": 157.3928680419922,
      "epoch": 0.08188153310104529,
      "grad_norm": 1.7190651893615723,
      "kl": 0.02734375,
      "learning_rate": 1e-06,
      "loss": 0.0011,
      "reward": 1.2142857909202576,
      "reward_std": 0.2253357619047165,
      "rewards/accuracy_reward": 0.267857164144516,
      "rewards/format_reward": 0.9464285969734192,
      "step": 47
    },
    {
      "completion_length": 115.51786422729492,
      "epoch": 0.08362369337979095,
      "grad_norm": 0.6601378321647644,
      "kl": 0.02496337890625,
      "learning_rate": 1e-06,
      "loss": 0.001,
      "reward": 1.1071429252624512,
      "reward_std": 0.0714285746216774,
      "rewards/accuracy_reward": 0.10714286006987095,
      "rewards/format_reward": 1.0,
      "step": 48
    },
    {
      "completion_length": 121.14286041259766,
      "epoch": 0.08536585365853659,
      "grad_norm": 1.0855175256729126,
      "kl": 0.025146484375,
      "learning_rate": 1e-06,
      "loss": 0.001,
      "reward": 1.2142857313156128,
      "reward_std": 0.0824786126613617,
      "rewards/accuracy_reward": 0.2142857313156128,
      "rewards/format_reward": 1.0,
      "step": 49
    },
    {
      "completion_length": 122.62500762939453,
      "epoch": 0.08710801393728224,
      "grad_norm": 1.4344055652618408,
      "kl": 0.0289306640625,
      "learning_rate": 1e-06,
      "loss": 0.0012,
      "reward": 1.0178571939468384,
      "reward_std": 0.0357142873108387,
      "rewards/accuracy_reward": 0.01785714365541935,
      "rewards/format_reward": 1.0,
      "step": 50
    },
    {
      "completion_length": 157.75000762939453,
      "epoch": 0.08885017421602788,
      "grad_norm": 0.6275267004966736,
      "kl": 0.016571044921875,
      "learning_rate": 1e-06,
      "loss": 0.0007,
      "reward": 1.3571429252624512,
      "reward_std": 0.1428571529686451,
      "rewards/accuracy_reward": 0.3571428805589676,
      "rewards/format_reward": 1.0,
      "step": 51
    },
    {
      "completion_length": 191.50000762939453,
      "epoch": 0.09059233449477352,
      "grad_norm": 0.9530377984046936,
      "kl": 0.0240478515625,
      "learning_rate": 1e-06,
      "loss": 0.001,
      "reward": 1.1785714626312256,
      "reward_std": 0.18409645557403564,
      "rewards/accuracy_reward": 0.19642857648432255,
      "rewards/format_reward": 0.9821428656578064,
      "step": 52
    },
    {
      "completion_length": 119.6785774230957,
      "epoch": 0.09233449477351917,
      "grad_norm": 1.568297266960144,
      "kl": 0.01593017578125,
      "learning_rate": 1e-06,
      "loss": 0.0006,
      "reward": 1.4821429252624512,
      "reward_std": 0.1785714402794838,
      "rewards/accuracy_reward": 0.4821428954601288,
      "rewards/format_reward": 1.0,
      "step": 53
    },
    {
      "completion_length": 192.0357208251953,
      "epoch": 0.09407665505226481,
      "grad_norm": 1.9954557418823242,
      "kl": 0.056884765625,
      "learning_rate": 1e-06,
      "loss": 0.0023,
      "reward": 1.4285714626312256,
      "reward_std": 0.2253357619047165,
      "rewards/accuracy_reward": 0.446428582072258,
      "rewards/format_reward": 0.9821428656578064,
      "step": 54
    },
    {
      "completion_length": 186.9821548461914,
      "epoch": 0.09581881533101046,
      "grad_norm": 3.1780290603637695,
      "kl": 0.0447998046875,
      "learning_rate": 1e-06,
      "loss": 0.0018,
      "reward": 1.2678571939468384,
      "reward_std": 0.25248411297798157,
      "rewards/accuracy_reward": 0.321428582072258,
      "rewards/format_reward": 0.9464285969734192,
      "step": 55
    },
    {
      "completion_length": 214.80358123779297,
      "epoch": 0.0975609756097561,
      "grad_norm": 1.200026273727417,
      "kl": 0.0296630859375,
      "learning_rate": 1e-06,
      "loss": 0.0012,
      "reward": 1.285714328289032,
      "reward_std": 0.2363857999444008,
      "rewards/accuracy_reward": 0.3035714477300644,
      "rewards/format_reward": 0.9821428656578064,
      "step": 56
    },
    {
      "completion_length": 141.85714721679688,
      "epoch": 0.09930313588850175,
      "grad_norm": 1.5381468534469604,
      "kl": 0.04248046875,
      "learning_rate": 1e-06,
      "loss": 0.0017,
      "reward": 1.3035715222358704,
      "reward_std": 0.14838216453790665,
      "rewards/accuracy_reward": 0.3035714477300644,
      "rewards/format_reward": 1.0,
      "step": 57
    },
    {
      "completion_length": 89.01786041259766,
      "epoch": 0.10104529616724739,
      "grad_norm": 3.5556106567382812,
      "kl": 0.03363037109375,
      "learning_rate": 1e-06,
      "loss": 0.0013,
      "reward": 1.4285714626312256,
      "reward_std": 0.2253357470035553,
      "rewards/accuracy_reward": 0.4285714477300644,
      "rewards/format_reward": 1.0,
      "step": 58
    },
    {
      "completion_length": 94.07143020629883,
      "epoch": 0.10278745644599303,
      "grad_norm": 0.8911753296852112,
      "kl": 0.0328369140625,
      "learning_rate": 1e-06,
      "loss": 0.0013,
      "reward": 1.2142857909202576,
      "reward_std": 0.0714285746216774,
      "rewards/accuracy_reward": 0.2321428656578064,
      "rewards/format_reward": 0.9821428656578064,
      "step": 59
    },
    {
      "completion_length": 113.00000762939453,
      "epoch": 0.10452961672473868,
      "grad_norm": 2.696057081222534,
      "kl": 0.0401611328125,
      "learning_rate": 1e-06,
      "loss": 0.0016,
      "reward": 1.160714328289032,
      "reward_std": 0.0357142873108387,
      "rewards/accuracy_reward": 0.160714291036129,
      "rewards/format_reward": 1.0,
      "step": 60
    },
    {
      "completion_length": 143.46429061889648,
      "epoch": 0.10627177700348432,
      "grad_norm": 2.277161121368408,
      "kl": 0.0289306640625,
      "learning_rate": 1e-06,
      "loss": 0.0012,
      "reward": 1.196428656578064,
      "reward_std": 0.1896214708685875,
      "rewards/accuracy_reward": 0.2142857201397419,
      "rewards/format_reward": 0.9821428656578064,
      "step": 61
    },
    {
      "completion_length": 41.250000953674316,
      "epoch": 0.10801393728222997,
      "grad_norm": 0.5966669321060181,
      "kl": 0.041748046875,
      "learning_rate": 1e-06,
      "loss": 0.0017,
      "reward": 1.410714328289032,
      "reward_std": 0.07695358991622925,
      "rewards/accuracy_reward": 0.4107143133878708,
      "rewards/format_reward": 1.0,
      "step": 62
    },
    {
      "completion_length": 106.0535774230957,
      "epoch": 0.10975609756097561,
      "grad_norm": 1.0405926704406738,
      "kl": 0.03521728515625,
      "learning_rate": 1e-06,
      "loss": 0.0014,
      "reward": 1.3214285969734192,
      "reward_std": 0.1428571529686451,
      "rewards/accuracy_reward": 0.321428582072258,
      "rewards/format_reward": 1.0,
      "step": 63
    },
    {
      "completion_length": 100.30357551574707,
      "epoch": 0.11149825783972125,
      "grad_norm": 2.6442556381225586,
      "kl": 0.043701171875,
      "learning_rate": 1e-06,
      "loss": 0.0017,
      "reward": 1.2142857313156128,
      "reward_std": 0.0714285746216774,
      "rewards/accuracy_reward": 0.2142857164144516,
      "rewards/format_reward": 1.0,
      "step": 64
    },
    {
      "completion_length": 127.28572463989258,
      "epoch": 0.1132404181184669,
      "grad_norm": 1.4323869943618774,
      "kl": 0.0272216796875,
      "learning_rate": 1e-06,
      "loss": 0.0011,
      "reward": 1.160714328289032,
      "reward_std": 0.13527478277683258,
      "rewards/accuracy_reward": 0.1964285746216774,
      "rewards/format_reward": 0.9642857313156128,
      "step": 65
    },
    {
      "completion_length": 91.3750057220459,
      "epoch": 0.11498257839721254,
      "grad_norm": 1.4564169645309448,
      "kl": 0.0450439453125,
      "learning_rate": 1e-06,
      "loss": 0.0018,
      "reward": 1.2500000596046448,
      "reward_std": 0.1539071835577488,
      "rewards/accuracy_reward": 0.267857164144516,
      "rewards/format_reward": 0.9821428656578064,
      "step": 66
    },
    {
      "completion_length": 110.78572082519531,
      "epoch": 0.11672473867595819,
      "grad_norm": 3.4178943634033203,
      "kl": 0.03350830078125,
      "learning_rate": 1e-06,
      "loss": 0.0013,
      "reward": 1.3392857909202576,
      "reward_std": 0.1785714402794838,
      "rewards/accuracy_reward": 0.3392857313156128,
      "rewards/format_reward": 1.0,
      "step": 67
    },
    {
      "completion_length": 54.58928871154785,
      "epoch": 0.11846689895470383,
      "grad_norm": 2.3461058139801025,
      "kl": 0.0557861328125,
      "learning_rate": 1e-06,
      "loss": 0.0022,
      "reward": 1.160714328289032,
      "reward_std": 0.1071428619325161,
      "rewards/accuracy_reward": 0.1607142984867096,
      "rewards/format_reward": 1.0,
      "step": 68
    },
    {
      "completion_length": 198.821439743042,
      "epoch": 0.12020905923344948,
      "grad_norm": 1.9061115980148315,
      "kl": 0.0479736328125,
      "learning_rate": 1e-06,
      "loss": 0.0019,
      "reward": 1.0892857909202576,
      "reward_std": 0.16546404361724854,
      "rewards/accuracy_reward": 0.1428571492433548,
      "rewards/format_reward": 0.9464285969734192,
      "step": 69
    },
    {
      "completion_length": 59.28571701049805,
      "epoch": 0.12195121951219512,
      "grad_norm": 2.6947081089019775,
      "kl": 0.0968017578125,
      "learning_rate": 1e-06,
      "loss": 0.0039,
      "reward": 1.1607143878936768,
      "reward_std": 0.1071428619325161,
      "rewards/accuracy_reward": 0.16071429289877415,
      "rewards/format_reward": 1.0,
      "step": 70
    },
    {
      "completion_length": 46.14285850524902,
      "epoch": 0.12369337979094076,
      "grad_norm": 3.665738582611084,
      "kl": 0.0404052734375,
      "learning_rate": 1e-06,
      "loss": 0.0016,
      "reward": 1.25,
      "reward_std": 0.18409645557403564,
      "rewards/accuracy_reward": 0.2678571492433548,
      "rewards/format_reward": 0.9821428656578064,
      "step": 71
    },
    {
      "completion_length": 49.37500190734863,
      "epoch": 0.1254355400696864,
      "grad_norm": 3.0653085708618164,
      "kl": 0.02880859375,
      "learning_rate": 1e-06,
      "loss": 0.0012,
      "reward": 1.2142857313156128,
      "reward_std": 0.11266788095235825,
      "rewards/accuracy_reward": 0.2142857313156128,
      "rewards/format_reward": 1.0,
      "step": 72
    },
    {
      "completion_length": 127.98215103149414,
      "epoch": 0.12717770034843207,
      "grad_norm": 0.9887677431106567,
      "kl": 0.03167724609375,
      "learning_rate": 1e-06,
      "loss": 0.0013,
      "reward": 1.2142857909202576,
      "reward_std": 0.0714285746216774,
      "rewards/accuracy_reward": 0.232142873108387,
      "rewards/format_reward": 0.9821428656578064,
      "step": 73
    },
    {
      "completion_length": 48.66071701049805,
      "epoch": 0.1289198606271777,
      "grad_norm": 0.8168498873710632,
      "kl": 0.03155517578125,
      "learning_rate": 1e-06,
      "loss": 0.0013,
      "reward": 1.160714328289032,
      "reward_std": 0.0357142873108387,
      "rewards/accuracy_reward": 0.160714291036129,
      "rewards/format_reward": 1.0,
      "step": 74
    },
    {
      "completion_length": 133.78571701049805,
      "epoch": 0.13066202090592335,
      "grad_norm": 0.4777332544326782,
      "kl": 0.03472900390625,
      "learning_rate": 1e-06,
      "loss": 0.0014,
      "reward": 1.1428571939468384,
      "reward_std": 0.0714285746216774,
      "rewards/accuracy_reward": 0.160714291036129,
      "rewards/format_reward": 0.9821428656578064,
      "step": 75
    },
    {
      "completion_length": 92.28572082519531,
      "epoch": 0.13240418118466898,
      "grad_norm": 1.1240160465240479,
      "kl": 0.03271484375,
      "learning_rate": 1e-06,
      "loss": 0.0013,
      "reward": 1.2678571939468384,
      "reward_std": 0.14838216453790665,
      "rewards/accuracy_reward": 0.2857142984867096,
      "rewards/format_reward": 0.9821428656578064,
      "step": 76
    },
    {
      "completion_length": 116.6785774230957,
      "epoch": 0.13414634146341464,
      "grad_norm": 2.4515292644500732,
      "kl": 0.044189453125,
      "learning_rate": 1e-06,
      "loss": 0.0018,
      "reward": 1.2678571939468384,
      "reward_std": 0.14838216453790665,
      "rewards/accuracy_reward": 0.2857142984867096,
      "rewards/format_reward": 0.9821428656578064,
      "step": 77
    },
    {
      "completion_length": 72.44643020629883,
      "epoch": 0.13588850174216027,
      "grad_norm": 2.802521228790283,
      "kl": 0.0426025390625,
      "learning_rate": 1e-06,
      "loss": 0.0017,
      "reward": 1.25,
      "reward_std": 0.1539071798324585,
      "rewards/accuracy_reward": 0.25,
      "rewards/format_reward": 1.0,
      "step": 78
    },
    {
      "completion_length": 105.0000057220459,
      "epoch": 0.13763066202090593,
      "grad_norm": 2.8100438117980957,
      "kl": 0.03338623046875,
      "learning_rate": 1e-06,
      "loss": 0.0013,
      "reward": 1.285714328289032,
      "reward_std": 0.24695907905697823,
      "rewards/accuracy_reward": 0.3035714477300644,
      "rewards/format_reward": 0.9821428656578064,
      "step": 79
    },
    {
      "completion_length": 52.67857360839844,
      "epoch": 0.13937282229965156,
      "grad_norm": 2.374569892883301,
      "kl": 0.0562744140625,
      "learning_rate": 1e-06,
      "loss": 0.0023,
      "reward": 1.3928571939468384,
      "reward_std": 0.18409645557403564,
      "rewards/accuracy_reward": 0.392857164144516,
      "rewards/format_reward": 1.0,
      "step": 80
    },
    {
      "completion_length": 40.535715103149414,
      "epoch": 0.14111498257839722,
      "grad_norm": 1.0314127206802368,
      "kl": 0.0440673828125,
      "learning_rate": 1e-06,
      "loss": 0.0018,
      "reward": 1.3571429252624512,
      "reward_std": 0.0714285746216774,
      "rewards/accuracy_reward": 0.35714287497103214,
      "rewards/format_reward": 1.0,
      "step": 81
    },
    {
      "completion_length": 81.91071701049805,
      "epoch": 0.14285714285714285,
      "grad_norm": 0.6449408531188965,
      "kl": 0.03765869140625,
      "learning_rate": 1e-06,
      "loss": 0.0015,
      "reward": 1.3214285969734192,
      "reward_std": 0.04123930633068085,
      "rewards/accuracy_reward": 0.3214285969734192,
      "rewards/format_reward": 1.0,
      "step": 82
    },
    {
      "completion_length": 111.01786422729492,
      "epoch": 0.1445993031358885,
      "grad_norm": 4.188468933105469,
      "kl": 0.033203125,
      "learning_rate": 1e-06,
      "loss": 0.0013,
      "reward": 1.321428656578064,
      "reward_std": 0.2142857238650322,
      "rewards/accuracy_reward": 0.3392857313156128,
      "rewards/format_reward": 0.9821428656578064,
      "step": 83
    },
    {
      "completion_length": 98.83929061889648,
      "epoch": 0.14634146341463414,
      "grad_norm": 2.2457661628723145,
      "kl": 0.03643798828125,
      "learning_rate": 1e-06,
      "loss": 0.0015,
      "reward": 1.410714328289032,
      "reward_std": 0.1896214708685875,
      "rewards/accuracy_reward": 0.4285714477300644,
      "rewards/format_reward": 0.9821428656578064,
      "step": 84
    },
    {
      "completion_length": 76.16071701049805,
      "epoch": 0.1480836236933798,
      "grad_norm": 1.101365566253662,
      "kl": 0.0364990234375,
      "learning_rate": 1e-06,
      "loss": 0.0015,
      "reward": 1.5535714626312256,
      "reward_std": 0.07695358991622925,
      "rewards/accuracy_reward": 0.5535714626312256,
      "rewards/format_reward": 1.0,
      "step": 85
    },
    {
      "completion_length": 97.35714721679688,
      "epoch": 0.14982578397212543,
      "grad_norm": 3.1208243370056152,
      "kl": 0.038818359375,
      "learning_rate": 1e-06,
      "loss": 0.0016,
      "reward": 1.2678571939468384,
      "reward_std": 0.1785714402794838,
      "rewards/accuracy_reward": 0.2857142984867096,
      "rewards/format_reward": 0.9821428656578064,
      "step": 86
    },
    {
      "completion_length": 42.07143020629883,
      "epoch": 0.15156794425087108,
      "grad_norm": 0.7317284941673279,
      "kl": 0.0546875,
      "learning_rate": 1e-06,
      "loss": 0.0022,
      "reward": 1.0714285969734192,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0714285746216774,
      "rewards/format_reward": 1.0,
      "step": 87
    },
    {
      "completion_length": 134.0178623199463,
      "epoch": 0.15331010452961671,
      "grad_norm": 2.284250259399414,
      "kl": 0.038818359375,
      "learning_rate": 1e-06,
      "loss": 0.0016,
      "reward": 1.1785714626312256,
      "reward_std": 0.17098908126354218,
      "rewards/accuracy_reward": 0.2142857238650322,
      "rewards/format_reward": 0.9642857313156128,
      "step": 88
    },
    {
      "completion_length": 71.32143020629883,
      "epoch": 0.15505226480836237,
      "grad_norm": 2.4166319370269775,
      "kl": 0.0408935546875,
      "learning_rate": 1e-06,
      "loss": 0.0016,
      "reward": 1.285714328289032,
      "reward_std": 0.18409644439816475,
      "rewards/accuracy_reward": 0.2857142984867096,
      "rewards/format_reward": 1.0,
      "step": 89
    },
    {
      "completion_length": 49.76785850524902,
      "epoch": 0.156794425087108,
      "grad_norm": 0.20406803488731384,
      "kl": 0.041015625,
      "learning_rate": 1e-06,
      "loss": 0.0016,
      "reward": 1.0714285969734192,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0714285746216774,
      "rewards/format_reward": 1.0,
      "step": 90
    },
    {
      "completion_length": 41.28571701049805,
      "epoch": 0.15853658536585366,
      "grad_norm": 1.4971178770065308,
      "kl": 0.0352783203125,
      "learning_rate": 1e-06,
      "loss": 0.0014,
      "reward": 1.6250001192092896,
      "reward_std": 0.07695358991622925,
      "rewards/accuracy_reward": 0.6250000298023224,
      "rewards/format_reward": 1.0,
      "step": 91
    },
    {
      "completion_length": 88.53572082519531,
      "epoch": 0.1602787456445993,
      "grad_norm": 3.6174209117889404,
      "kl": 0.03338623046875,
      "learning_rate": 1e-06,
      "loss": 0.0013,
      "reward": 1.4464285969734192,
      "reward_std": 0.23086077719926834,
      "rewards/accuracy_reward": 0.4464285969734192,
      "rewards/format_reward": 1.0,
      "step": 92
    },
    {
      "completion_length": 88.9464340209961,
      "epoch": 0.16202090592334495,
      "grad_norm": 0.5585963726043701,
      "kl": 0.0479736328125,
      "learning_rate": 1e-06,
      "loss": 0.0019,
      "reward": 1.2678571939468384,
      "reward_std": 0.0357142873108387,
      "rewards/accuracy_reward": 0.2857142984867096,
      "rewards/format_reward": 0.9821428656578064,
      "step": 93
    },
    {
      "completion_length": 100.58929061889648,
      "epoch": 0.16376306620209058,
      "grad_norm": 0.45368415117263794,
      "kl": 0.0296630859375,
      "learning_rate": 1e-06,
      "loss": 0.0012,
      "reward": 1.3392857909202576,
      "reward_std": 0.1071428619325161,
      "rewards/accuracy_reward": 0.3392857313156128,
      "rewards/format_reward": 1.0,
      "step": 94
    },
    {
      "completion_length": 77.55357360839844,
      "epoch": 0.16550522648083624,
      "grad_norm": 1.1148852109909058,
      "kl": 0.03955078125,
      "learning_rate": 1e-06,
      "loss": 0.0016,
      "reward": 1.2500000596046448,
      "reward_std": 0.11266787722706795,
      "rewards/accuracy_reward": 0.2500000149011612,
      "rewards/format_reward": 1.0,
      "step": 95
    },
    {
      "completion_length": 87.23214721679688,
      "epoch": 0.1672473867595819,
      "grad_norm": 2.378681182861328,
      "kl": 0.0572509765625,
      "learning_rate": 1e-06,
      "loss": 0.0023,
      "reward": 1.3035715222358704,
      "reward_std": 0.14838216453790665,
      "rewards/accuracy_reward": 0.3035714477300644,
      "rewards/format_reward": 1.0,
      "step": 96
    },
    {
      "completion_length": 78.42857551574707,
      "epoch": 0.16898954703832753,
      "grad_norm": 2.708867073059082,
      "kl": 0.0322265625,
      "learning_rate": 1e-06,
      "loss": 0.0013,
      "reward": 1.1785715222358704,
      "reward_std": 0.0714285746216774,
      "rewards/accuracy_reward": 0.17857144214212894,
      "rewards/format_reward": 1.0,
      "step": 97
    },
    {
      "completion_length": 91.42857360839844,
      "epoch": 0.17073170731707318,
      "grad_norm": 1.7530436515808105,
      "kl": 0.04833984375,
      "learning_rate": 1e-06,
      "loss": 0.0019,
      "reward": 1.1964285969734192,
      "reward_std": 0.07695358991622925,
      "rewards/accuracy_reward": 0.1964285746216774,
      "rewards/format_reward": 1.0,
      "step": 98
    },
    {
      "completion_length": 120.48214721679688,
      "epoch": 0.17247386759581881,
      "grad_norm": 2.45546293258667,
      "kl": 0.03021240234375,
      "learning_rate": 1e-06,
      "loss": 0.0012,
      "reward": 1.5000000596046448,
      "reward_std": 0.25552502274513245,
      "rewards/accuracy_reward": 0.517857164144516,
      "rewards/format_reward": 0.9821428656578064,
      "step": 99
    },
    {
      "completion_length": 189.16072845458984,
      "epoch": 0.17421602787456447,
      "grad_norm": 1.9118529558181763,
      "kl": 0.0391845703125,
      "learning_rate": 1e-06,
      "loss": 0.0016,
      "reward": 1.196428656578064,
      "reward_std": 0.15086627006530762,
      "rewards/accuracy_reward": 0.2321428656578064,
      "rewards/format_reward": 0.9642857313156128,
      "step": 100
    },
    {
      "completion_length": 74.30357551574707,
      "epoch": 0.1759581881533101,
      "grad_norm": 3.851543664932251,
      "kl": 0.0513916015625,
      "learning_rate": 1e-06,
      "loss": 0.0021,
      "reward": 1.3214285969734192,
      "reward_std": 0.2967643216252327,
      "rewards/accuracy_reward": 0.3214285969734192,
      "rewards/format_reward": 1.0,
      "step": 101
    },
    {
      "completion_length": 119.73214721679688,
      "epoch": 0.17770034843205576,
      "grad_norm": 1.3045148849487305,
      "kl": 0.040771484375,
      "learning_rate": 1e-06,
      "loss": 0.0016,
      "reward": 1.3928571939468384,
      "reward_std": 0.0714285746216774,
      "rewards/accuracy_reward": 0.392857164144516,
      "rewards/format_reward": 1.0,
      "step": 102
    },
    {
      "completion_length": 97.01786041259766,
      "epoch": 0.1794425087108014,
      "grad_norm": 1.8413246870040894,
      "kl": 0.02679443359375,
      "learning_rate": 1e-06,
      "loss": 0.0011,
      "reward": 1.3392857909202576,
      "reward_std": 0.07695358991622925,
      "rewards/accuracy_reward": 0.3392857313156128,
      "rewards/format_reward": 1.0,
      "step": 103
    },
    {
      "completion_length": 66.01785850524902,
      "epoch": 0.18118466898954705,
      "grad_norm": 1.7732882499694824,
      "kl": 0.0533447265625,
      "learning_rate": 1e-06,
      "loss": 0.0021,
      "reward": 1.3035714626312256,
      "reward_std": 0.14838216826319695,
      "rewards/accuracy_reward": 0.3035714402794838,
      "rewards/format_reward": 1.0,
      "step": 104
    },
    {
      "completion_length": 106.39286041259766,
      "epoch": 0.18292682926829268,
      "grad_norm": 1.3419760465621948,
      "kl": 0.026611328125,
      "learning_rate": 1e-06,
      "loss": 0.0011,
      "reward": 1.446428656578064,
      "reward_std": 0.14838217198848724,
      "rewards/accuracy_reward": 0.446428582072258,
      "rewards/format_reward": 1.0,
      "step": 105
    },
    {
      "completion_length": 158.4464340209961,
      "epoch": 0.18466898954703834,
      "grad_norm": 1.9505287408828735,
      "kl": 0.03173828125,
      "learning_rate": 1e-06,
      "loss": 0.0013,
      "reward": 1.2142857313156128,
      "reward_std": 0.1539071798324585,
      "rewards/accuracy_reward": 0.2321428656578064,
      "rewards/format_reward": 0.9821428656578064,
      "step": 106
    },
    {
      "completion_length": 92.5714340209961,
      "epoch": 0.18641114982578397,
      "grad_norm": 2.1574547290802,
      "kl": 0.0264892578125,
      "learning_rate": 1e-06,
      "loss": 0.0011,
      "reward": 1.3214285969734192,
      "reward_std": 0.11266788095235825,
      "rewards/accuracy_reward": 0.3214285969734192,
      "rewards/format_reward": 1.0,
      "step": 107
    },
    {
      "completion_length": 91.50000381469727,
      "epoch": 0.18815331010452963,
      "grad_norm": 0.6726558804512024,
      "kl": 0.03436279296875,
      "learning_rate": 1e-06,
      "loss": 0.0014,
      "reward": 1.321428656578064,
      "reward_std": 0.0714285746216774,
      "rewards/accuracy_reward": 0.3392857313156128,
      "rewards/format_reward": 0.9821428656578064,
      "step": 108
    },
    {
      "completion_length": 72.64286041259766,
      "epoch": 0.18989547038327526,
      "grad_norm": 2.206634521484375,
      "kl": 0.0567626953125,
      "learning_rate": 1e-06,
      "loss": 0.0023,
      "reward": 1.4821429252624512,
      "reward_std": 0.1181928999722004,
      "rewards/accuracy_reward": 0.4821428805589676,
      "rewards/format_reward": 1.0,
      "step": 109
    },
    {
      "completion_length": 156.51786422729492,
      "epoch": 0.1916376306620209,
      "grad_norm": 2.635490655899048,
      "kl": 0.047607421875,
      "learning_rate": 1e-06,
      "loss": 0.0019,
      "reward": 1.4285715222358704,
      "reward_std": 0.1649572253227234,
      "rewards/accuracy_reward": 0.4285714477300644,
      "rewards/format_reward": 1.0,
      "step": 110
    },
    {
      "completion_length": 72.64286041259766,
      "epoch": 0.19337979094076654,
      "grad_norm": 3.589315414428711,
      "kl": 0.043212890625,
      "learning_rate": 1e-06,
      "loss": 0.0017,
      "reward": 1.3928572535514832,
      "reward_std": 0.1428571492433548,
      "rewards/accuracy_reward": 0.3928571492433548,
      "rewards/format_reward": 1.0,
      "step": 111
    },
    {
      "completion_length": 139.69643783569336,
      "epoch": 0.1951219512195122,
      "grad_norm": 0.8147983551025391,
      "kl": 0.0255126953125,
      "learning_rate": 1e-06,
      "loss": 0.001,
      "reward": 1.285714328289032,
      "reward_std": 0.0714285746216774,
      "rewards/accuracy_reward": 0.2857142984867096,
      "rewards/format_reward": 1.0,
      "step": 112
    },
    {
      "completion_length": 173.71429443359375,
      "epoch": 0.19686411149825783,
      "grad_norm": 3.0332417488098145,
      "kl": 0.03369140625,
      "learning_rate": 1e-06,
      "loss": 0.0013,
      "reward": 1.0892857313156128,
      "reward_std": 0.14838215708732605,
      "rewards/accuracy_reward": 0.1071428656578064,
      "rewards/format_reward": 0.9821428656578064,
      "step": 113
    },
    {
      "completion_length": 204.16072845458984,
      "epoch": 0.1986062717770035,
      "grad_norm": 2.643592357635498,
      "kl": 0.1168212890625,
      "learning_rate": 1e-06,
      "loss": 0.0047,
      "reward": 1.1071429252624512,
      "reward_std": 0.16242313012480736,
      "rewards/accuracy_reward": 0.1428571492433548,
      "rewards/format_reward": 0.9642857313156128,
      "step": 114
    },
    {
      "completion_length": 110.07143020629883,
      "epoch": 0.20034843205574912,
      "grad_norm": 0.8254410624504089,
      "kl": 0.036376953125,
      "learning_rate": 1e-06,
      "loss": 0.0015,
      "reward": 1.1428571939468384,
      "reward_std": 0.0714285746216774,
      "rewards/accuracy_reward": 0.160714291036129,
      "rewards/format_reward": 0.9821428656578064,
      "step": 115
    },
    {
      "completion_length": 95.58929061889648,
      "epoch": 0.20209059233449478,
      "grad_norm": 1.296014666557312,
      "kl": 0.0499267578125,
      "learning_rate": 1e-06,
      "loss": 0.002,
      "reward": 1.3214285969734192,
      "reward_std": 0.0714285746216774,
      "rewards/accuracy_reward": 0.321428582072258,
      "rewards/format_reward": 1.0,
      "step": 116
    },
    {
      "completion_length": 68.51786041259766,
      "epoch": 0.2038327526132404,
      "grad_norm": 2.4811439514160156,
      "kl": 0.0330810546875,
      "learning_rate": 1e-06,
      "loss": 0.0013,
      "reward": 1.3928571939468384,
      "reward_std": 0.0714285746216774,
      "rewards/accuracy_reward": 0.392857164144516,
      "rewards/format_reward": 1.0,
      "step": 117
    },
    {
      "completion_length": 96.14286041259766,
      "epoch": 0.20557491289198607,
      "grad_norm": 0.46663737297058105,
      "kl": 0.0428466796875,
      "learning_rate": 1e-06,
      "loss": 0.0017,
      "reward": 1.3928572535514832,
      "reward_std": 0.0714285746216774,
      "rewards/accuracy_reward": 0.3928571492433548,
      "rewards/format_reward": 1.0,
      "step": 118
    },
    {
      "completion_length": 83.26786422729492,
      "epoch": 0.2073170731707317,
      "grad_norm": 1.1690462827682495,
      "kl": 0.037353515625,
      "learning_rate": 1e-06,
      "loss": 0.0015,
      "reward": 1.3035715222358704,
      "reward_std": 0.0357142873108387,
      "rewards/accuracy_reward": 0.3035714477300644,
      "rewards/format_reward": 1.0,
      "step": 119
    },
    {
      "completion_length": 133.21428680419922,
      "epoch": 0.20905923344947736,
      "grad_norm": 2.5085010528564453,
      "kl": 0.05029296875,
      "learning_rate": 1e-06,
      "loss": 0.002,
      "reward": 1.2142857909202576,
      "reward_std": 0.18409644439816475,
      "rewards/accuracy_reward": 0.23214287497103214,
      "rewards/format_reward": 0.9821428656578064,
      "step": 120
    },
    {
      "completion_length": 143.7857208251953,
      "epoch": 0.21080139372822299,
      "grad_norm": 4.637099742889404,
      "kl": 0.045166015625,
      "learning_rate": 1e-06,
      "loss": 0.0018,
      "reward": 1.285714328289032,
      "reward_std": 0.1539071798324585,
      "rewards/accuracy_reward": 0.3035714477300644,
      "rewards/format_reward": 0.9821428656578064,
      "step": 121
    },
    {
      "completion_length": 64.19643211364746,
      "epoch": 0.21254355400696864,
      "grad_norm": 1.975265622138977,
      "kl": 0.03314208984375,
      "learning_rate": 1e-06,
      "loss": 0.0013,
      "reward": 1.2321428656578064,
      "reward_std": 0.14838216453790665,
      "rewards/accuracy_reward": 0.2321428656578064,
      "rewards/format_reward": 1.0,
      "step": 122
    },
    {
      "completion_length": 91.57143211364746,
      "epoch": 0.21428571428571427,
      "grad_norm": 2.885680913925171,
      "kl": 0.072509765625,
      "learning_rate": 1e-06,
      "loss": 0.0029,
      "reward": 1.2500000596046448,
      "reward_std": 0.1539071872830391,
      "rewards/accuracy_reward": 0.2500000149011612,
      "rewards/format_reward": 1.0,
      "step": 123
    },
    {
      "completion_length": 118.39286041259766,
      "epoch": 0.21602787456445993,
      "grad_norm": 1.6002825498580933,
      "kl": 0.0328369140625,
      "learning_rate": 1e-06,
      "loss": 0.0013,
      "reward": 1.1250000596046448,
      "reward_std": 0.1071428619325161,
      "rewards/accuracy_reward": 0.160714291036129,
      "rewards/format_reward": 0.9642857313156128,
      "step": 124
    },
    {
      "completion_length": 86.53571891784668,
      "epoch": 0.21777003484320556,
      "grad_norm": 1.9014854431152344,
      "kl": 0.0528564453125,
      "learning_rate": 1e-06,
      "loss": 0.0021,
      "reward": 1.2678571939468384,
      "reward_std": 0.14838216453790665,
      "rewards/accuracy_reward": 0.267857164144516,
      "rewards/format_reward": 1.0,
      "step": 125
    },
    {
      "completion_length": 76.4464340209961,
      "epoch": 0.21951219512195122,
      "grad_norm": 1.3376573324203491,
      "kl": 0.02325439453125,
      "learning_rate": 1e-06,
      "loss": 0.0009,
      "reward": 1.3571429252624512,
      "reward_std": 0.18409645557403564,
      "rewards/accuracy_reward": 0.3571428656578064,
      "rewards/format_reward": 1.0,
      "step": 126
    },
    {
      "completion_length": 73.57143020629883,
      "epoch": 0.22125435540069685,
      "grad_norm": 1.281449794769287,
      "kl": 0.0526123046875,
      "learning_rate": 1e-06,
      "loss": 0.0021,
      "reward": 1.0535714626312256,
      "reward_std": 0.07695358991622925,
      "rewards/accuracy_reward": 0.0535714328289032,
      "rewards/format_reward": 1.0,
      "step": 127
    },
    {
      "completion_length": 75.46429061889648,
      "epoch": 0.2229965156794425,
      "grad_norm": 1.3834271430969238,
      "kl": 0.0509033203125,
      "learning_rate": 1e-06,
      "loss": 0.002,
      "reward": 1.160714328289032,
      "reward_std": 0.0357142873108387,
      "rewards/accuracy_reward": 0.160714291036129,
      "rewards/format_reward": 1.0,
      "step": 128
    },
    {
      "completion_length": 83.48214721679688,
      "epoch": 0.22473867595818817,
      "grad_norm": 3.221709966659546,
      "kl": 0.048583984375,
      "learning_rate": 1e-06,
      "loss": 0.0019,
      "reward": 1.2678571939468384,
      "reward_std": 0.29123931378126144,
      "rewards/accuracy_reward": 0.2857142984867096,
      "rewards/format_reward": 0.9821428656578064,
      "step": 129
    },
    {
      "completion_length": 82.48214530944824,
      "epoch": 0.2264808362369338,
      "grad_norm": 2.775517463684082,
      "kl": 0.046630859375,
      "learning_rate": 1e-06,
      "loss": 0.0019,
      "reward": 1.4642857313156128,
      "reward_std": 0.11266787722706795,
      "rewards/accuracy_reward": 0.4642857313156128,
      "rewards/format_reward": 1.0,
      "step": 130
    },
    {
      "completion_length": 106.03571701049805,
      "epoch": 0.22822299651567945,
      "grad_norm": 0.7142723798751831,
      "kl": 0.0277099609375,
      "learning_rate": 1e-06,
      "loss": 0.0011,
      "reward": 1.2500000596046448,
      "reward_std": 0.11266787722706795,
      "rewards/accuracy_reward": 0.2500000149011612,
      "rewards/format_reward": 1.0,
      "step": 131
    },
    {
      "completion_length": 54.875003814697266,
      "epoch": 0.22996515679442509,
      "grad_norm": 1.956363320350647,
      "kl": 0.0419921875,
      "learning_rate": 1e-06,
      "loss": 0.0017,
      "reward": 1.2321428656578064,
      "reward_std": 0.07695359364151955,
      "rewards/accuracy_reward": 0.2321428656578064,
      "rewards/format_reward": 1.0,
      "step": 132
    },
    {
      "completion_length": 133.0357208251953,
      "epoch": 0.23170731707317074,
      "grad_norm": 2.0702030658721924,
      "kl": 0.0458984375,
      "learning_rate": 1e-06,
      "loss": 0.0018,
      "reward": 1.3035714626312256,
      "reward_std": 0.2610500380396843,
      "rewards/accuracy_reward": 0.3392857313156128,
      "rewards/format_reward": 0.9642857313156128,
      "step": 133
    },
    {
      "completion_length": 123.44643783569336,
      "epoch": 0.23344947735191637,
      "grad_norm": 3.096449136734009,
      "kl": 0.0501708984375,
      "learning_rate": 1e-06,
      "loss": 0.002,
      "reward": 1.2678571939468384,
      "reward_std": 0.14838217198848724,
      "rewards/accuracy_reward": 0.267857164144516,
      "rewards/format_reward": 1.0,
      "step": 134
    },
    {
      "completion_length": 129.83929061889648,
      "epoch": 0.23519163763066203,
      "grad_norm": 3.0806472301483154,
      "kl": 0.03887939453125,
      "learning_rate": 1e-06,
      "loss": 0.0016,
      "reward": 1.3392857909202576,
      "reward_std": 0.1785714365541935,
      "rewards/accuracy_reward": 0.3571428805589676,
      "rewards/format_reward": 0.9821428656578064,
      "step": 135
    },
    {
      "completion_length": 134.85715103149414,
      "epoch": 0.23693379790940766,
      "grad_norm": 2.5985164642333984,
      "kl": 0.04425048828125,
      "learning_rate": 1e-06,
      "loss": 0.0018,
      "reward": 1.4285714626312256,
      "reward_std": 0.25552502274513245,
      "rewards/accuracy_reward": 0.4285714328289032,
      "rewards/format_reward": 1.0,
      "step": 136
    },
    {
      "completion_length": 122.83929061889648,
      "epoch": 0.23867595818815332,
      "grad_norm": 0.6726248860359192,
      "kl": 0.06396484375,
      "learning_rate": 1e-06,
      "loss": 0.0026,
      "reward": 1.3571429252624512,
      "reward_std": 0.0714285746216774,
      "rewards/accuracy_reward": 0.3571428805589676,
      "rewards/format_reward": 1.0,
      "step": 137
    },
    {
      "completion_length": 93.76786041259766,
      "epoch": 0.24041811846689895,
      "grad_norm": 1.511077880859375,
      "kl": 0.02423095703125,
      "learning_rate": 1e-06,
      "loss": 0.001,
      "reward": 1.2857143878936768,
      "reward_std": 0.0714285746216774,
      "rewards/accuracy_reward": 0.2857142984867096,
      "rewards/format_reward": 1.0,
      "step": 138
    },
    {
      "completion_length": 98.94643020629883,
      "epoch": 0.2421602787456446,
      "grad_norm": 1.0339641571044922,
      "kl": 0.0565185546875,
      "learning_rate": 1e-06,
      "loss": 0.0023,
      "reward": 1.071428656578064,
      "reward_std": 0.0714285746216774,
      "rewards/accuracy_reward": 0.07142857648432255,
      "rewards/format_reward": 1.0,
      "step": 139
    },
    {
      "completion_length": 124.57143592834473,
      "epoch": 0.24390243902439024,
      "grad_norm": 1.6461817026138306,
      "kl": 0.0247802734375,
      "learning_rate": 1e-06,
      "loss": 0.001,
      "reward": 1.3750000596046448,
      "reward_std": 0.21981074661016464,
      "rewards/accuracy_reward": 0.392857164144516,
      "rewards/format_reward": 0.9821428656578064,
      "step": 140
    },
    {
      "completion_length": 109.5714340209961,
      "epoch": 0.2456445993031359,
      "grad_norm": 1.4544490575790405,
      "kl": 0.042236328125,
      "learning_rate": 1e-06,
      "loss": 0.0017,
      "reward": 1.2321429252624512,
      "reward_std": 0.1071428656578064,
      "rewards/accuracy_reward": 0.2500000149011612,
      "rewards/format_reward": 0.9821428656578064,
      "step": 141
    },
    {
      "completion_length": 154.23215293884277,
      "epoch": 0.24738675958188153,
      "grad_norm": 4.0668230056762695,
      "kl": 0.02862548828125,
      "learning_rate": 1e-06,
      "loss": 0.0011,
      "reward": 1.196428656578064,
      "reward_std": 0.28061603009700775,
      "rewards/accuracy_reward": 0.2321428656578064,
      "rewards/format_reward": 0.9642857313156128,
      "step": 142
    },
    {
      "completion_length": 110.10715103149414,
      "epoch": 0.24912891986062718,
      "grad_norm": 3.707798957824707,
      "kl": 0.0501708984375,
      "learning_rate": 1e-06,
      "loss": 0.002,
      "reward": 1.4107143878936768,
      "reward_std": 0.1181928962469101,
      "rewards/accuracy_reward": 0.4107142984867096,
      "rewards/format_reward": 1.0,
      "step": 143
    },
    {
      "completion_length": 151.73214721679688,
      "epoch": 0.2508710801393728,
      "grad_norm": 1.6972342729568481,
      "kl": 0.041259765625,
      "learning_rate": 1e-06,
      "loss": 0.0016,
      "reward": 1.2500000596046448,
      "reward_std": 0.20117833092808723,
      "rewards/accuracy_reward": 0.2857142984867096,
      "rewards/format_reward": 0.9642857313156128,
      "step": 144
    },
    {
      "completion_length": 101.66072082519531,
      "epoch": 0.25261324041811845,
      "grad_norm": 3.099278211593628,
      "kl": 0.141845703125,
      "learning_rate": 1e-06,
      "loss": 0.0057,
      "reward": 1.1785714626312256,
      "reward_std": 0.1428571529686451,
      "rewards/accuracy_reward": 0.1785714328289032,
      "rewards/format_reward": 1.0,
      "step": 145
    },
    {
      "completion_length": 211.41071701049805,
      "epoch": 0.25435540069686413,
      "grad_norm": 1.7294187545776367,
      "kl": 0.031005859375,
      "learning_rate": 1e-06,
      "loss": 0.0012,
      "reward": 1.2321429252624512,
      "reward_std": 0.1896214783191681,
      "rewards/accuracy_reward": 0.2857143059372902,
      "rewards/format_reward": 0.9464285969734192,
      "step": 146
    },
    {
      "completion_length": 53.32143020629883,
      "epoch": 0.25609756097560976,
      "grad_norm": 2.51180100440979,
      "kl": 0.032470703125,
      "learning_rate": 1e-06,
      "loss": 0.0013,
      "reward": 1.196428656578064,
      "reward_std": 0.1071428619325161,
      "rewards/accuracy_reward": 0.2142857238650322,
      "rewards/format_reward": 0.9821428656578064,
      "step": 147
    },
    {
      "completion_length": 144.10715103149414,
      "epoch": 0.2578397212543554,
      "grad_norm": 1.118396282196045,
      "kl": 0.03857421875,
      "learning_rate": 1e-06,
      "loss": 0.0015,
      "reward": 1.160714328289032,
      "reward_std": 0.14838217198848724,
      "rewards/accuracy_reward": 0.2142857164144516,
      "rewards/format_reward": 0.9464285969734192,
      "step": 148
    },
    {
      "completion_length": 209.16072845458984,
      "epoch": 0.259581881533101,
      "grad_norm": 0.5762439966201782,
      "kl": 0.023193359375,
      "learning_rate": 1e-06,
      "loss": 0.0009,
      "reward": 1.1785714626312256,
      "reward_std": 0.0714285746216774,
      "rewards/accuracy_reward": 0.2142857238650322,
      "rewards/format_reward": 0.9642857313156128,
      "step": 149
    },
    {
      "completion_length": 180.80358123779297,
      "epoch": 0.2613240418118467,
      "grad_norm": 1.6064391136169434,
      "kl": 0.064453125,
      "learning_rate": 1e-06,
      "loss": 0.0026,
      "reward": 1.1964285969734192,
      "reward_std": 0.13527477905154228,
      "rewards/accuracy_reward": 0.2142857164144516,
      "rewards/format_reward": 0.9821428656578064,
      "step": 150
    },
    {
      "completion_length": 140.00000762939453,
      "epoch": 0.26306620209059234,
      "grad_norm": 2.3106741905212402,
      "kl": 0.0313720703125,
      "learning_rate": 1e-06,
      "loss": 0.0013,
      "reward": 1.3750000596046448,
      "reward_std": 0.07695359364151955,
      "rewards/accuracy_reward": 0.392857164144516,
      "rewards/format_reward": 0.9821428656578064,
      "step": 151
    },
    {
      "completion_length": 137.25000381469727,
      "epoch": 0.26480836236933797,
      "grad_norm": 1.3025450706481934,
      "kl": 0.02471923828125,
      "learning_rate": 1e-06,
      "loss": 0.001,
      "reward": 1.1785714626312256,
      "reward_std": 0.0714285746216774,
      "rewards/accuracy_reward": 0.1964285746216774,
      "rewards/format_reward": 0.9821428656578064,
      "step": 152
    },
    {
      "completion_length": 139.60715103149414,
      "epoch": 0.2665505226480836,
      "grad_norm": 2.006934881210327,
      "kl": 0.039306640625,
      "learning_rate": 1e-06,
      "loss": 0.0016,
      "reward": 1.2321429252624512,
      "reward_std": 0.1896214671432972,
      "rewards/accuracy_reward": 0.2321428656578064,
      "rewards/format_reward": 1.0,
      "step": 153
    },
    {
      "completion_length": 146.92857360839844,
      "epoch": 0.2682926829268293,
      "grad_norm": 2.0233981609344482,
      "kl": 0.0511474609375,
      "learning_rate": 1e-06,
      "loss": 0.002,
      "reward": 1.3750000596046448,
      "reward_std": 0.2610500380396843,
      "rewards/accuracy_reward": 0.3928571715950966,
      "rewards/format_reward": 0.9821428656578064,
      "step": 154
    },
    {
      "completion_length": 135.76786422729492,
      "epoch": 0.2700348432055749,
      "grad_norm": 2.7513368129730225,
      "kl": 0.03076171875,
      "learning_rate": 1e-06,
      "loss": 0.0012,
      "reward": 1.2321429252624512,
      "reward_std": 0.1181928962469101,
      "rewards/accuracy_reward": 0.232142873108387,
      "rewards/format_reward": 1.0,
      "step": 155
    },
    {
      "completion_length": 135.73214721679688,
      "epoch": 0.27177700348432055,
      "grad_norm": 2.0850443840026855,
      "kl": 0.0506591796875,
      "learning_rate": 1e-06,
      "loss": 0.002,
      "reward": 1.196428656578064,
      "reward_std": 0.1071428656578064,
      "rewards/accuracy_reward": 0.2142857238650322,
      "rewards/format_reward": 0.9821428656578064,
      "step": 156
    },
    {
      "completion_length": 153.05358123779297,
      "epoch": 0.2735191637630662,
      "grad_norm": 2.3688433170318604,
      "kl": 0.022216796875,
      "learning_rate": 1e-06,
      "loss": 0.0009,
      "reward": 1.6250000596046448,
      "reward_std": 0.14838216826319695,
      "rewards/accuracy_reward": 0.6428571939468384,
      "rewards/format_reward": 0.9821428656578064,
      "step": 157
    },
    {
      "completion_length": 205.92858123779297,
      "epoch": 0.27526132404181186,
      "grad_norm": 3.015334129333496,
      "kl": 0.03314208984375,
      "learning_rate": 1e-06,
      "loss": 0.0013,
      "reward": 1.2500000596046448,
      "reward_std": 0.26657508313655853,
      "rewards/accuracy_reward": 0.2857142984867096,
      "rewards/format_reward": 0.9642857313156128,
      "step": 158
    },
    {
      "completion_length": 167.4107208251953,
      "epoch": 0.2770034843205575,
      "grad_norm": 1.3877893686294556,
      "kl": 0.035400390625,
      "learning_rate": 1e-06,
      "loss": 0.0014,
      "reward": 1.1250000596046448,
      "reward_std": 0.1071428619325161,
      "rewards/accuracy_reward": 0.1428571492433548,
      "rewards/format_reward": 0.9821428656578064,
      "step": 159
    },
    {
      "completion_length": 135.0357208251953,
      "epoch": 0.2787456445993031,
      "grad_norm": 2.2060704231262207,
      "kl": 0.030029296875,
      "learning_rate": 1e-06,
      "loss": 0.0012,
      "reward": 1.285714328289032,
      "reward_std": 0.24241763353347778,
      "rewards/accuracy_reward": 0.3035714477300644,
      "rewards/format_reward": 0.9821428656578064,
      "step": 160
    },
    {
      "completion_length": 185.50000762939453,
      "epoch": 0.2804878048780488,
      "grad_norm": 0.6886353492736816,
      "kl": 0.032958984375,
      "learning_rate": 1e-06,
      "loss": 0.0013,
      "reward": 1.160714328289032,
      "reward_std": 0.1071428619325161,
      "rewards/accuracy_reward": 0.1785714328289032,
      "rewards/format_reward": 0.9821428656578064,
      "step": 161
    },
    {
      "completion_length": 84.0535774230957,
      "epoch": 0.28222996515679444,
      "grad_norm": 1.9014931917190552,
      "kl": 0.0433349609375,
      "learning_rate": 1e-06,
      "loss": 0.0017,
      "reward": 1.3571429252624512,
      "reward_std": 0.18409645557403564,
      "rewards/accuracy_reward": 0.3571428805589676,
      "rewards/format_reward": 1.0,
      "step": 162
    },
    {
      "completion_length": 95.14286041259766,
      "epoch": 0.28397212543554007,
      "grad_norm": 5.170721054077148,
      "kl": 0.0498046875,
      "learning_rate": 1e-06,
      "loss": 0.002,
      "reward": 1.4821429252624512,
      "reward_std": 0.1071428656578064,
      "rewards/accuracy_reward": 0.4821428805589676,
      "rewards/format_reward": 1.0,
      "step": 163
    },
    {
      "completion_length": 120.8214340209961,
      "epoch": 0.2857142857142857,
      "grad_norm": 4.6296796798706055,
      "kl": 0.035888671875,
      "learning_rate": 1e-06,
      "loss": 0.0014,
      "reward": 1.0535714626312256,
      "reward_std": 0.1071428656578064,
      "rewards/accuracy_reward": 0.0714285746216774,
      "rewards/format_reward": 0.9821428656578064,
      "step": 164
    },
    {
      "completion_length": 68.62500190734863,
      "epoch": 0.2874564459930314,
      "grad_norm": 2.599226236343384,
      "kl": 0.0472412109375,
      "learning_rate": 1e-06,
      "loss": 0.0019,
      "reward": 1.3571429252624512,
      "reward_std": 0.19514648616313934,
      "rewards/accuracy_reward": 0.3571428805589676,
      "rewards/format_reward": 1.0,
      "step": 165
    },
    {
      "completion_length": 114.92858123779297,
      "epoch": 0.289198606271777,
      "grad_norm": 0.9972745776176453,
      "kl": 0.0364990234375,
      "learning_rate": 1e-06,
      "loss": 0.0015,
      "reward": 1.285714328289032,
      "reward_std": 0.0714285746216774,
      "rewards/accuracy_reward": 0.3035714477300644,
      "rewards/format_reward": 0.9821428656578064,
      "step": 166
    },
    {
      "completion_length": 132.44643020629883,
      "epoch": 0.29094076655052264,
      "grad_norm": 3.7945616245269775,
      "kl": 0.03155517578125,
      "learning_rate": 1e-06,
      "loss": 0.0013,
      "reward": 1.160714328289032,
      "reward_std": 0.1785714402794838,
      "rewards/accuracy_reward": 0.1607142984867096,
      "rewards/format_reward": 1.0,
      "step": 167
    },
    {
      "completion_length": 65.51786041259766,
      "epoch": 0.2926829268292683,
      "grad_norm": 2.3588600158691406,
      "kl": 0.0594482421875,
      "learning_rate": 1e-06,
      "loss": 0.0024,
      "reward": 1.3750000596046448,
      "reward_std": 0.1181928962469101,
      "rewards/accuracy_reward": 0.3750000149011612,
      "rewards/format_reward": 1.0,
      "step": 168
    },
    {
      "completion_length": 97.87500762939453,
      "epoch": 0.29442508710801396,
      "grad_norm": 1.6221623420715332,
      "kl": 0.0416259765625,
      "learning_rate": 1e-06,
      "loss": 0.0017,
      "reward": 1.3392857909202576,
      "reward_std": 0.0357142873108387,
      "rewards/accuracy_reward": 0.3392857313156128,
      "rewards/format_reward": 1.0,
      "step": 169
    },
    {
      "completion_length": 68.78571510314941,
      "epoch": 0.2961672473867596,
      "grad_norm": 3.4078965187072754,
      "kl": 0.197509765625,
      "learning_rate": 1e-06,
      "loss": 0.0079,
      "reward": 1.321428656578064,
      "reward_std": 0.1428571492433548,
      "rewards/accuracy_reward": 0.321428582072258,
      "rewards/format_reward": 1.0,
      "step": 170
    },
    {
      "completion_length": 95.67857360839844,
      "epoch": 0.2979094076655052,
      "grad_norm": 3.5229647159576416,
      "kl": 0.03277587890625,
      "learning_rate": 1e-06,
      "loss": 0.0013,
      "reward": 1.2142857313156128,
      "reward_std": 0.1539071835577488,
      "rewards/accuracy_reward": 0.2142857164144516,
      "rewards/format_reward": 1.0,
      "step": 171
    },
    {
      "completion_length": 92.17857360839844,
      "epoch": 0.29965156794425085,
      "grad_norm": 1.6305819749832153,
      "kl": 0.0592041015625,
      "learning_rate": 1e-06,
      "loss": 0.0024,
      "reward": 1.25,
      "reward_std": 0.1539071798324585,
      "rewards/accuracy_reward": 0.25,
      "rewards/format_reward": 1.0,
      "step": 172
    },
    {
      "completion_length": 134.91072273254395,
      "epoch": 0.30139372822299654,
      "grad_norm": 0.8862522840499878,
      "kl": 0.03143310546875,
      "learning_rate": 1e-06,
      "loss": 0.0013,
      "reward": 1.3928571939468384,
      "reward_std": 0.0714285746216774,
      "rewards/accuracy_reward": 0.392857164144516,
      "rewards/format_reward": 1.0,
      "step": 173
    },
    {
      "completion_length": 91.64286041259766,
      "epoch": 0.30313588850174217,
      "grad_norm": 2.307363748550415,
      "kl": 0.04156494140625,
      "learning_rate": 1e-06,
      "loss": 0.0017,
      "reward": 1.1785715222358704,
      "reward_std": 0.18409645557403564,
      "rewards/accuracy_reward": 0.1785714365541935,
      "rewards/format_reward": 1.0,
      "step": 174
    },
    {
      "completion_length": 100.71429061889648,
      "epoch": 0.3048780487804878,
      "grad_norm": 4.026115417480469,
      "kl": 0.03656005859375,
      "learning_rate": 1e-06,
      "loss": 0.0015,
      "reward": 1.410714328289032,
      "reward_std": 0.1785714402794838,
      "rewards/accuracy_reward": 0.4285714328289032,
      "rewards/format_reward": 0.9821428656578064,
      "step": 175
    },
    {
      "completion_length": 122.00000762939453,
      "epoch": 0.30662020905923343,
      "grad_norm": 1.8479260206222534,
      "kl": 0.04150390625,
      "learning_rate": 1e-06,
      "loss": 0.0017,
      "reward": 1.446428656578064,
      "reward_std": 0.1071428619325161,
      "rewards/accuracy_reward": 0.4464285969734192,
      "rewards/format_reward": 1.0,
      "step": 176
    },
    {
      "completion_length": 99.00000381469727,
      "epoch": 0.3083623693379791,
      "grad_norm": 2.3721132278442383,
      "kl": 0.0333251953125,
      "learning_rate": 1e-06,
      "loss": 0.0013,
      "reward": 1.3571428656578064,
      "reward_std": 0.11266787722706795,
      "rewards/accuracy_reward": 0.3571428656578064,
      "rewards/format_reward": 1.0,
      "step": 177
    },
    {
      "completion_length": 132.9285774230957,
      "epoch": 0.31010452961672474,
      "grad_norm": 0.8678131103515625,
      "kl": 0.03631591796875,
      "learning_rate": 1e-06,
      "loss": 0.0015,
      "reward": 1.3214285969734192,
      "reward_std": 0.18409644439816475,
      "rewards/accuracy_reward": 0.321428582072258,
      "rewards/format_reward": 1.0,
      "step": 178
    },
    {
      "completion_length": 112.21428680419922,
      "epoch": 0.3118466898954704,
      "grad_norm": 2.4448330402374268,
      "kl": 0.043212890625,
      "learning_rate": 1e-06,
      "loss": 0.0017,
      "reward": 1.160714328289032,
      "reward_std": 0.1181928962469101,
      "rewards/accuracy_reward": 0.1607142984867096,
      "rewards/format_reward": 1.0,
      "step": 179
    },
    {
      "completion_length": 134.64286041259766,
      "epoch": 0.313588850174216,
      "grad_norm": 1.2125186920166016,
      "kl": 0.0335693359375,
      "learning_rate": 1e-06,
      "loss": 0.0013,
      "reward": 1.2500000596046448,
      "reward_std": 0.0714285746216774,
      "rewards/accuracy_reward": 0.2500000149011612,
      "rewards/format_reward": 1.0,
      "step": 180
    },
    {
      "completion_length": 174.6428680419922,
      "epoch": 0.3153310104529617,
      "grad_norm": 1.7917309999465942,
      "kl": 0.025146484375,
      "learning_rate": 1e-06,
      "loss": 0.001,
      "reward": 1.2321429252624512,
      "reward_std": 0.14838216826319695,
      "rewards/accuracy_reward": 0.2500000149011612,
      "rewards/format_reward": 0.9821428656578064,
      "step": 181
    },
    {
      "completion_length": 203.71429443359375,
      "epoch": 0.3170731707317073,
      "grad_norm": 3.1097638607025146,
      "kl": 0.03955078125,
      "learning_rate": 1e-06,
      "loss": 0.0016,
      "reward": 1.3392857909202576,
      "reward_std": 0.18105553090572357,
      "rewards/accuracy_reward": 0.3571428805589676,
      "rewards/format_reward": 0.9821428656578064,
      "step": 182
    },
    {
      "completion_length": 145.0357208251953,
      "epoch": 0.31881533101045295,
      "grad_norm": 0.9235899448394775,
      "kl": 0.0269775390625,
      "learning_rate": 1e-06,
      "loss": 0.0011,
      "reward": 1.3035715222358704,
      "reward_std": 0.1785714402794838,
      "rewards/accuracy_reward": 0.3035714477300644,
      "rewards/format_reward": 1.0,
      "step": 183
    },
    {
      "completion_length": 143.55358123779297,
      "epoch": 0.3205574912891986,
      "grad_norm": 0.8848127126693726,
      "kl": 0.02520751953125,
      "learning_rate": 1e-06,
      "loss": 0.001,
      "reward": 1.2500000596046448,
      "reward_std": 0.1428571529686451,
      "rewards/accuracy_reward": 0.25000000931322575,
      "rewards/format_reward": 1.0,
      "step": 184
    },
    {
      "completion_length": 165.62500762939453,
      "epoch": 0.32229965156794427,
      "grad_norm": 1.5377029180526733,
      "kl": 0.017181396484375,
      "learning_rate": 1e-06,
      "loss": 0.0007,
      "reward": 1.2321428656578064,
      "reward_std": 0.14838216453790665,
      "rewards/accuracy_reward": 0.2321428656578064,
      "rewards/format_reward": 1.0,
      "step": 185
    },
    {
      "completion_length": 156.3928680419922,
      "epoch": 0.3240418118466899,
      "grad_norm": 1.4587819576263428,
      "kl": 0.02685546875,
      "learning_rate": 1e-06,
      "loss": 0.0011,
      "reward": 1.3928571939468384,
      "reward_std": 0.19514650478959084,
      "rewards/accuracy_reward": 0.392857164144516,
      "rewards/format_reward": 1.0,
      "step": 186
    },
    {
      "completion_length": 216.4107208251953,
      "epoch": 0.32578397212543553,
      "grad_norm": 2.4346601963043213,
      "kl": 0.018524169921875,
      "learning_rate": 1e-06,
      "loss": 0.0007,
      "reward": 1.410714328289032,
      "reward_std": 0.14838216453790665,
      "rewards/accuracy_reward": 0.4107143133878708,
      "rewards/format_reward": 1.0,
      "step": 187
    },
    {
      "completion_length": 161.73215103149414,
      "epoch": 0.32752613240418116,
      "grad_norm": 0.7127730250358582,
      "kl": 0.02191162109375,
      "learning_rate": 1e-06,
      "loss": 0.0009,
      "reward": 1.2321429252624512,
      "reward_std": 0.1071428619325161,
      "rewards/accuracy_reward": 0.2321428693830967,
      "rewards/format_reward": 1.0,
      "step": 188
    },
    {
      "completion_length": 217.33929443359375,
      "epoch": 0.32926829268292684,
      "grad_norm": 1.9412953853607178,
      "kl": 0.012176513671875,
      "learning_rate": 1e-06,
      "loss": 0.0005,
      "reward": 1.2321429252624512,
      "reward_std": 0.1785714402794838,
      "rewards/accuracy_reward": 0.2321428656578064,
      "rewards/format_reward": 1.0,
      "step": 189
    },
    {
      "completion_length": 315.4464416503906,
      "epoch": 0.3310104529616725,
      "grad_norm": 0.8375619053840637,
      "kl": 0.01995849609375,
      "learning_rate": 1e-06,
      "loss": 0.0008,
      "reward": 1.2321429252624512,
      "reward_std": 0.21124479547142982,
      "rewards/accuracy_reward": 0.2500000149011612,
      "rewards/format_reward": 0.9821428656578064,
      "step": 190
    },
    {
      "completion_length": 203.0178680419922,
      "epoch": 0.3327526132404181,
      "grad_norm": 2.4890501499176025,
      "kl": 0.0247802734375,
      "learning_rate": 1e-06,
      "loss": 0.001,
      "reward": 1.2857143878936768,
      "reward_std": 0.18409645557403564,
      "rewards/accuracy_reward": 0.2857142984867096,
      "rewards/format_reward": 1.0,
      "step": 191
    },
    {
      "completion_length": 217.75000762939453,
      "epoch": 0.3344947735191638,
      "grad_norm": 0.957855761051178,
      "kl": 0.014923095703125,
      "learning_rate": 1e-06,
      "loss": 0.0006,
      "reward": 1.5714285969734192,
      "reward_std": 0.25552502274513245,
      "rewards/accuracy_reward": 0.5714285969734192,
      "rewards/format_reward": 1.0,
      "step": 192
    },
    {
      "completion_length": 347.08929443359375,
      "epoch": 0.3362369337979094,
      "grad_norm": 0.6019410490989685,
      "kl": 0.0164794921875,
      "learning_rate": 1e-06,
      "loss": 0.0007,
      "reward": 1.3035714626312256,
      "reward_std": 0.18105553090572357,
      "rewards/accuracy_reward": 0.321428582072258,
      "rewards/format_reward": 0.9821428656578064,
      "step": 193
    },
    {
      "completion_length": 268.8214416503906,
      "epoch": 0.33797909407665505,
      "grad_norm": 0.979367733001709,
      "kl": 0.02947998046875,
      "learning_rate": 1e-06,
      "loss": 0.0012,
      "reward": 1.3928571939468384,
      "reward_std": 0.26657505333423615,
      "rewards/accuracy_reward": 0.4107143133878708,
      "rewards/format_reward": 0.9821428656578064,
      "step": 194
    },
    {
      "completion_length": 332.1785888671875,
      "epoch": 0.3397212543554007,
      "grad_norm": 1.4663631916046143,
      "kl": 0.0245361328125,
      "learning_rate": 1e-06,
      "loss": 0.001,
      "reward": 1.3214285969734192,
      "reward_std": 0.3571428656578064,
      "rewards/accuracy_reward": 0.392857164144516,
      "rewards/format_reward": 0.9285714626312256,
      "step": 195
    },
    {
      "completion_length": 389.5357360839844,
      "epoch": 0.34146341463414637,
      "grad_norm": 0.43615275621414185,
      "kl": 0.016143798828125,
      "learning_rate": 1e-06,
      "loss": 0.0006,
      "reward": 1.1250000596046448,
      "reward_std": 0.1785714402794838,
      "rewards/accuracy_reward": 0.1785714402794838,
      "rewards/format_reward": 0.9464285969734192,
      "step": 196
    },
    {
      "completion_length": 267.7678680419922,
      "epoch": 0.343205574912892,
      "grad_norm": 2.834510564804077,
      "kl": 0.0343017578125,
      "learning_rate": 1e-06,
      "loss": 0.0014,
      "reward": 1.3392857313156128,
      "reward_std": 0.14838216826319695,
      "rewards/accuracy_reward": 0.3571428656578064,
      "rewards/format_reward": 0.9821428656578064,
      "step": 197
    },
    {
      "completion_length": 261.1607208251953,
      "epoch": 0.34494773519163763,
      "grad_norm": 1.6161222457885742,
      "kl": 0.0169677734375,
      "learning_rate": 1e-06,
      "loss": 0.0007,
      "reward": 1.5000000596046448,
      "reward_std": 0.26657505333423615,
      "rewards/accuracy_reward": 0.517857164144516,
      "rewards/format_reward": 0.9821428656578064,
      "step": 198
    },
    {
      "completion_length": 414.0357360839844,
      "epoch": 0.34668989547038326,
      "grad_norm": 1.6385655403137207,
      "kl": 0.0218505859375,
      "learning_rate": 1e-06,
      "loss": 0.0009,
      "reward": 1.410714328289032,
      "reward_std": 0.29123930633068085,
      "rewards/accuracy_reward": 0.464285746216774,
      "rewards/format_reward": 0.9464285969734192,
      "step": 199
    },
    {
      "completion_length": 352.7678680419922,
      "epoch": 0.34843205574912894,
      "grad_norm": 1.2244716882705688,
      "kl": 0.01873779296875,
      "learning_rate": 1e-06,
      "loss": 0.0008,
      "reward": 1.071428656578064,
      "reward_std": 0.24241764843463898,
      "rewards/accuracy_reward": 0.1250000074505806,
      "rewards/format_reward": 0.9464285969734192,
      "step": 200
    },
    {
      "completion_length": 327.78572845458984,
      "epoch": 0.3501742160278746,
      "grad_norm": 0.8990840315818787,
      "kl": 0.02777099609375,
      "learning_rate": 1e-06,
      "loss": 0.0011,
      "reward": 1.1964285969734192,
      "reward_std": 0.23086076974868774,
      "rewards/accuracy_reward": 0.2500000149011612,
      "rewards/format_reward": 0.9464285969734192,
      "step": 201
    },
    {
      "completion_length": 249.0535888671875,
      "epoch": 0.3519163763066202,
      "grad_norm": 0.8288552165031433,
      "kl": 0.026885986328125,
      "learning_rate": 1e-06,
      "loss": 0.0011,
      "reward": 1.3035715222358704,
      "reward_std": 0.1071428619325161,
      "rewards/accuracy_reward": 0.3035714477300644,
      "rewards/format_reward": 1.0,
      "step": 202
    },
    {
      "completion_length": 291.83929443359375,
      "epoch": 0.35365853658536583,
      "grad_norm": 0.9214010238647461,
      "kl": 0.0103759765625,
      "learning_rate": 1e-06,
      "loss": 0.0004,
      "reward": 1.3750000596046448,
      "reward_std": 0.2891819700598717,
      "rewards/accuracy_reward": 0.3928571492433548,
      "rewards/format_reward": 0.9821428656578064,
      "step": 203
    },
    {
      "completion_length": 197.3928680419922,
      "epoch": 0.3554006968641115,
      "grad_norm": 1.698442816734314,
      "kl": 0.0186767578125,
      "learning_rate": 1e-06,
      "loss": 0.0007,
      "reward": 1.321428656578064,
      "reward_std": 0.1539071798324585,
      "rewards/accuracy_reward": 0.3392857313156128,
      "rewards/format_reward": 0.9821428656578064,
      "step": 204
    },
    {
      "completion_length": 191.39286041259766,
      "epoch": 0.35714285714285715,
      "grad_norm": 1.9810231924057007,
      "kl": 0.0213623046875,
      "learning_rate": 1e-06,
      "loss": 0.0009,
      "reward": 1.4464285969734192,
      "reward_std": 0.23086077719926834,
      "rewards/accuracy_reward": 0.464285746216774,
      "rewards/format_reward": 0.9821428656578064,
      "step": 205
    },
    {
      "completion_length": 208.1071548461914,
      "epoch": 0.3588850174216028,
      "grad_norm": 1.2316064834594727,
      "kl": 0.0518798828125,
      "learning_rate": 1e-06,
      "loss": 0.0021,
      "reward": 1.0892857909202576,
      "reward_std": 0.14838216826319695,
      "rewards/accuracy_reward": 0.08928571827709675,
      "rewards/format_reward": 1.0,
      "step": 206
    },
    {
      "completion_length": 213.3571548461914,
      "epoch": 0.3606271777003484,
      "grad_norm": 0.23318666219711304,
      "kl": 0.02264404296875,
      "learning_rate": 1e-06,
      "loss": 0.0009,
      "reward": 1.2500000596046448,
      "reward_std": 0.0714285746216774,
      "rewards/accuracy_reward": 0.2678571492433548,
      "rewards/format_reward": 0.9821428656578064,
      "step": 207
    },
    {
      "completion_length": 224.3571548461914,
      "epoch": 0.3623693379790941,
      "grad_norm": 1.5259240865707397,
      "kl": 0.01904296875,
      "learning_rate": 1e-06,
      "loss": 0.0008,
      "reward": 1.3571429252624512,
      "reward_std": 0.11266788095235825,
      "rewards/accuracy_reward": 0.3571428656578064,
      "rewards/format_reward": 1.0,
      "step": 208
    },
    {
      "completion_length": 217.94644165039062,
      "epoch": 0.3641114982578397,
      "grad_norm": 1.5131481885910034,
      "kl": 0.017578125,
      "learning_rate": 1e-06,
      "loss": 0.0007,
      "reward": 1.3214285969734192,
      "reward_std": 0.1428571529686451,
      "rewards/accuracy_reward": 0.3392857313156128,
      "rewards/format_reward": 0.9821428656578064,
      "step": 209
    },
    {
      "completion_length": 206.46429443359375,
      "epoch": 0.36585365853658536,
      "grad_norm": 0.8935233950614929,
      "kl": 0.025146484375,
      "learning_rate": 1e-06,
      "loss": 0.001,
      "reward": 1.1785715222358704,
      "reward_std": 0.19514649361371994,
      "rewards/accuracy_reward": 0.196428582072258,
      "rewards/format_reward": 0.9821428656578064,
      "step": 210
    },
    {
      "completion_length": 150.60714721679688,
      "epoch": 0.367595818815331,
      "grad_norm": 1.949971318244934,
      "kl": 0.108154296875,
      "learning_rate": 1e-06,
      "loss": 0.0043,
      "reward": 1.2857143878936768,
      "reward_std": 0.18409645557403564,
      "rewards/accuracy_reward": 0.3035714477300644,
      "rewards/format_reward": 0.9821428656578064,
      "step": 211
    },
    {
      "completion_length": 233.19644165039062,
      "epoch": 0.3693379790940767,
      "grad_norm": 1.5521166324615479,
      "kl": 0.0205078125,
      "learning_rate": 1e-06,
      "loss": 0.0008,
      "reward": 1.2678571939468384,
      "reward_std": 0.14838216826319695,
      "rewards/accuracy_reward": 0.2857142984867096,
      "rewards/format_reward": 0.9821428656578064,
      "step": 212
    },
    {
      "completion_length": 115.98214721679688,
      "epoch": 0.3710801393728223,
      "grad_norm": 2.789050579071045,
      "kl": 0.0338134765625,
      "learning_rate": 1e-06,
      "loss": 0.0013,
      "reward": 1.3035715222358704,
      "reward_std": 0.1071428656578064,
      "rewards/accuracy_reward": 0.3035714477300644,
      "rewards/format_reward": 1.0,
      "step": 213
    },
    {
      "completion_length": 148.96429061889648,
      "epoch": 0.37282229965156793,
      "grad_norm": 3.145935297012329,
      "kl": 0.02362060546875,
      "learning_rate": 1e-06,
      "loss": 0.0009,
      "reward": 1.2678572535514832,
      "reward_std": 0.21981074661016464,
      "rewards/accuracy_reward": 0.3035714328289032,
      "rewards/format_reward": 0.9642857313156128,
      "step": 214
    },
    {
      "completion_length": 129.25000762939453,
      "epoch": 0.37456445993031356,
      "grad_norm": 4.017449378967285,
      "kl": 0.0260009765625,
      "learning_rate": 1e-06,
      "loss": 0.001,
      "reward": 1.3392857909202576,
      "reward_std": 0.1785714402794838,
      "rewards/accuracy_reward": 0.3392857313156128,
      "rewards/format_reward": 1.0,
      "step": 215
    },
    {
      "completion_length": 204.2678680419922,
      "epoch": 0.37630662020905925,
      "grad_norm": 0.5732056498527527,
      "kl": 0.0830078125,
      "learning_rate": 1e-06,
      "loss": 0.0033,
      "reward": 1.2857143878936768,
      "reward_std": 0.0714285746216774,
      "rewards/accuracy_reward": 0.2857142984867096,
      "rewards/format_reward": 1.0,
      "step": 216
    },
    {
      "completion_length": 181.7678680419922,
      "epoch": 0.3780487804878049,
      "grad_norm": 0.7782284617424011,
      "kl": 0.068359375,
      "learning_rate": 1e-06,
      "loss": 0.0027,
      "reward": 1.3035714626312256,
      "reward_std": 0.07695358991622925,
      "rewards/accuracy_reward": 0.3214285746216774,
      "rewards/format_reward": 0.9821428656578064,
      "step": 217
    },
    {
      "completion_length": 142.875,
      "epoch": 0.3797909407665505,
      "grad_norm": 1.5171810388565063,
      "kl": 0.0218505859375,
      "learning_rate": 1e-06,
      "loss": 0.0009,
      "reward": 1.2678571939468384,
      "reward_std": 0.07695359364151955,
      "rewards/accuracy_reward": 0.267857164144516,
      "rewards/format_reward": 1.0,
      "step": 218
    },
    {
      "completion_length": 174.3214340209961,
      "epoch": 0.38153310104529614,
      "grad_norm": 1.3412084579467773,
      "kl": 0.13818359375,
      "learning_rate": 1e-06,
      "loss": 0.0055,
      "reward": 1.3392857313156128,
      "reward_std": 0.14838216826319695,
      "rewards/accuracy_reward": 0.3571428656578064,
      "rewards/format_reward": 0.9821428656578064,
      "step": 219
    },
    {
      "completion_length": 166.33929443359375,
      "epoch": 0.3832752613240418,
      "grad_norm": 2.023908853530884,
      "kl": 0.038818359375,
      "learning_rate": 1e-06,
      "loss": 0.0016,
      "reward": 1.2500000596046448,
      "reward_std": 0.18409645557403564,
      "rewards/accuracy_reward": 0.267857164144516,
      "rewards/format_reward": 0.9821428656578064,
      "step": 220
    },
    {
      "completion_length": 104.26786422729492,
      "epoch": 0.38501742160278746,
      "grad_norm": 0.8557510375976562,
      "kl": 0.02642822265625,
      "learning_rate": 1e-06,
      "loss": 0.0011,
      "reward": 1.5000000596046448,
      "reward_std": 0.11266788095235825,
      "rewards/accuracy_reward": 0.5000000298023224,
      "rewards/format_reward": 1.0,
      "step": 221
    },
    {
      "completion_length": 116.12500762939453,
      "epoch": 0.3867595818815331,
      "grad_norm": 2.272390365600586,
      "kl": 0.0286865234375,
      "learning_rate": 1e-06,
      "loss": 0.0011,
      "reward": 1.321428656578064,
      "reward_std": 0.0714285746216774,
      "rewards/accuracy_reward": 0.3214285746216774,
      "rewards/format_reward": 1.0,
      "step": 222
    },
    {
      "completion_length": 134.21429443359375,
      "epoch": 0.3885017421602787,
      "grad_norm": 2.2785820960998535,
      "kl": 0.05206298828125,
      "learning_rate": 1e-06,
      "loss": 0.0021,
      "reward": 1.2321429252624512,
      "reward_std": 0.1181928999722004,
      "rewards/accuracy_reward": 0.2321428656578064,
      "rewards/format_reward": 1.0,
      "step": 223
    },
    {
      "completion_length": 103.01786041259766,
      "epoch": 0.3902439024390244,
      "grad_norm": 1.0930382013320923,
      "kl": 0.022918701171875,
      "learning_rate": 1e-06,
      "loss": 0.0009,
      "reward": 1.4821429252624512,
      "reward_std": 0.14838216453790665,
      "rewards/accuracy_reward": 0.4821428805589676,
      "rewards/format_reward": 1.0,
      "step": 224
    },
    {
      "completion_length": 164.50000762939453,
      "epoch": 0.39198606271777003,
      "grad_norm": 1.6921836137771606,
      "kl": 0.02490234375,
      "learning_rate": 1e-06,
      "loss": 0.001,
      "reward": 1.2500000596046448,
      "reward_std": 0.18409644439816475,
      "rewards/accuracy_reward": 0.2500000149011612,
      "rewards/format_reward": 1.0,
      "step": 225
    },
    {
      "completion_length": 112.5535774230957,
      "epoch": 0.39372822299651566,
      "grad_norm": 2.7419986724853516,
      "kl": 0.04638671875,
      "learning_rate": 1e-06,
      "loss": 0.0019,
      "reward": 1.3035714626312256,
      "reward_std": 0.1896214671432972,
      "rewards/accuracy_reward": 0.3035714477300644,
      "rewards/format_reward": 1.0,
      "step": 226
    },
    {
      "completion_length": 127.62500381469727,
      "epoch": 0.39547038327526135,
      "grad_norm": 6.331331253051758,
      "kl": 0.03118896484375,
      "learning_rate": 1e-06,
      "loss": 0.0012,
      "reward": 1.285714328289032,
      "reward_std": 0.18409645557403564,
      "rewards/accuracy_reward": 0.3035714477300644,
      "rewards/format_reward": 0.9821428656578064,
      "step": 227
    },
    {
      "completion_length": 172.58929443359375,
      "epoch": 0.397212543554007,
      "grad_norm": 2.7327375411987305,
      "kl": 0.03802490234375,
      "learning_rate": 1e-06,
      "loss": 0.0015,
      "reward": 1.3928571939468384,
      "reward_std": 0.2253357470035553,
      "rewards/accuracy_reward": 0.4107142984867096,
      "rewards/format_reward": 0.9821428656578064,
      "step": 228
    },
    {
      "completion_length": 151.58929443359375,
      "epoch": 0.3989547038327526,
      "grad_norm": 0.866677463054657,
      "kl": 0.01922607421875,
      "learning_rate": 1e-06,
      "loss": 0.0008,
      "reward": 1.5178572535514832,
      "reward_std": 0.1181928962469101,
      "rewards/accuracy_reward": 0.5178571790456772,
      "rewards/format_reward": 1.0,
      "step": 229
    },
    {
      "completion_length": 175.2857208251953,
      "epoch": 0.40069686411149824,
      "grad_norm": 1.3600616455078125,
      "kl": 0.024658203125,
      "learning_rate": 1e-06,
      "loss": 0.001,
      "reward": 1.2142857313156128,
      "reward_std": 0.0714285746216774,
      "rewards/accuracy_reward": 0.23214287497103214,
      "rewards/format_reward": 0.9821428656578064,
      "step": 230
    },
    {
      "completion_length": 107.00000762939453,
      "epoch": 0.4024390243902439,
      "grad_norm": 3.2532806396484375,
      "kl": 0.04315185546875,
      "learning_rate": 1e-06,
      "loss": 0.0017,
      "reward": 1.5178571939468384,
      "reward_std": 0.4149572402238846,
      "rewards/accuracy_reward": 0.5357143133878708,
      "rewards/format_reward": 0.9821428656578064,
      "step": 231
    },
    {
      "completion_length": 183.3928680419922,
      "epoch": 0.40418118466898956,
      "grad_norm": 2.559502601623535,
      "kl": 0.03631591796875,
      "learning_rate": 1e-06,
      "loss": 0.0014,
      "reward": 1.321428656578064,
      "reward_std": 0.25552502274513245,
      "rewards/accuracy_reward": 0.3392857201397419,
      "rewards/format_reward": 0.9821428656578064,
      "step": 232
    },
    {
      "completion_length": 110.0714340209961,
      "epoch": 0.4059233449477352,
      "grad_norm": 2.014815330505371,
      "kl": 0.035888671875,
      "learning_rate": 1e-06,
      "loss": 0.0014,
      "reward": 1.410714328289032,
      "reward_std": 0.07695359364151955,
      "rewards/accuracy_reward": 0.4107142984867096,
      "rewards/format_reward": 1.0,
      "step": 233
    },
    {
      "completion_length": 164.33929443359375,
      "epoch": 0.4076655052264808,
      "grad_norm": 1.9719363451004028,
      "kl": 0.0830078125,
      "learning_rate": 1e-06,
      "loss": 0.0033,
      "reward": 1.2678571939468384,
      "reward_std": 0.2610500529408455,
      "rewards/accuracy_reward": 0.2857142984867096,
      "rewards/format_reward": 0.9821428656578064,
      "step": 234
    },
    {
      "completion_length": 167.98214721679688,
      "epoch": 0.4094076655052265,
      "grad_norm": 1.2324802875518799,
      "kl": 0.04840087890625,
      "learning_rate": 1e-06,
      "loss": 0.0019,
      "reward": 1.3750000596046448,
      "reward_std": 0.1071428619325161,
      "rewards/accuracy_reward": 0.3750000149011612,
      "rewards/format_reward": 1.0,
      "step": 235
    },
    {
      "completion_length": 130.76786041259766,
      "epoch": 0.41114982578397213,
      "grad_norm": 2.3901517391204834,
      "kl": 0.0355224609375,
      "learning_rate": 1e-06,
      "loss": 0.0014,
      "reward": 1.2142857313156128,
      "reward_std": 0.1428571529686451,
      "rewards/accuracy_reward": 0.2142857164144516,
      "rewards/format_reward": 1.0,
      "step": 236
    },
    {
      "completion_length": 82.35714530944824,
      "epoch": 0.41289198606271776,
      "grad_norm": 1.5934151411056519,
      "kl": 0.06884765625,
      "learning_rate": 1e-06,
      "loss": 0.0027,
      "reward": 1.160714328289032,
      "reward_std": 0.0357142873108387,
      "rewards/accuracy_reward": 0.160714291036129,
      "rewards/format_reward": 1.0,
      "step": 237
    },
    {
      "completion_length": 101.0535774230957,
      "epoch": 0.4146341463414634,
      "grad_norm": 1.6057491302490234,
      "kl": 0.081787109375,
      "learning_rate": 1e-06,
      "loss": 0.0033,
      "reward": 1.1428571939468384,
      "reward_std": 0.0714285746216774,
      "rewards/accuracy_reward": 0.1428571492433548,
      "rewards/format_reward": 1.0,
      "step": 238
    },
    {
      "completion_length": 134.33929443359375,
      "epoch": 0.4163763066202091,
      "grad_norm": 11.595154762268066,
      "kl": 0.03668212890625,
      "learning_rate": 1e-06,
      "loss": 0.0015,
      "reward": 1.3392857909202576,
      "reward_std": 0.16546405106782913,
      "rewards/accuracy_reward": 0.3750000298023224,
      "rewards/format_reward": 0.9642857313156128,
      "step": 239
    },
    {
      "completion_length": 116.76786422729492,
      "epoch": 0.4181184668989547,
      "grad_norm": 0.3346155881881714,
      "kl": 0.034912109375,
      "learning_rate": 1e-06,
      "loss": 0.0014,
      "reward": 1.0892857909202576,
      "reward_std": 0.0357142873108387,
      "rewards/accuracy_reward": 0.08928571827709675,
      "rewards/format_reward": 1.0,
      "step": 240
    },
    {
      "completion_length": 99.01786041259766,
      "epoch": 0.41986062717770034,
      "grad_norm": 1.3594080209732056,
      "kl": 0.0355224609375,
      "learning_rate": 1e-06,
      "loss": 0.0014,
      "reward": 1.3035714626312256,
      "reward_std": 0.1071428656578064,
      "rewards/accuracy_reward": 0.3035714328289032,
      "rewards/format_reward": 1.0,
      "step": 241
    },
    {
      "completion_length": 118.0714340209961,
      "epoch": 0.42160278745644597,
      "grad_norm": 1.0064547061920166,
      "kl": 0.0494384765625,
      "learning_rate": 1e-06,
      "loss": 0.002,
      "reward": 1.3392857909202576,
      "reward_std": 0.07695359364151955,
      "rewards/accuracy_reward": 0.3571428805589676,
      "rewards/format_reward": 0.9821428656578064,
      "step": 242
    },
    {
      "completion_length": 72.0714340209961,
      "epoch": 0.42334494773519166,
      "grad_norm": 0.7986550331115723,
      "kl": 0.0435791015625,
      "learning_rate": 1e-06,
      "loss": 0.0017,
      "reward": 1.3214285969734192,
      "reward_std": 0.04123930633068085,
      "rewards/accuracy_reward": 0.3214285969734192,
      "rewards/format_reward": 1.0,
      "step": 243
    },
    {
      "completion_length": 109.64286041259766,
      "epoch": 0.4250871080139373,
      "grad_norm": 1.521716833114624,
      "kl": 0.06268310546875,
      "learning_rate": 1e-06,
      "loss": 0.0025,
      "reward": 1.3750000596046448,
      "reward_std": 0.1071428619325161,
      "rewards/accuracy_reward": 0.3750000298023224,
      "rewards/format_reward": 1.0,
      "step": 244
    },
    {
      "completion_length": 64.33928871154785,
      "epoch": 0.4268292682926829,
      "grad_norm": 1.0075122117996216,
      "kl": 0.037353515625,
      "learning_rate": 1e-06,
      "loss": 0.0015,
      "reward": 1.4285715222358704,
      "reward_std": 0.0714285746216774,
      "rewards/accuracy_reward": 0.4285714477300644,
      "rewards/format_reward": 1.0,
      "step": 245
    },
    {
      "completion_length": 134.35715103149414,
      "epoch": 0.42857142857142855,
      "grad_norm": 1.5240962505340576,
      "kl": 0.03082275390625,
      "learning_rate": 1e-06,
      "loss": 0.0012,
      "reward": 1.2321428656578064,
      "reward_std": 0.1181928962469101,
      "rewards/accuracy_reward": 0.2321428656578064,
      "rewards/format_reward": 1.0,
      "step": 246
    },
    {
      "completion_length": 201.75000762939453,
      "epoch": 0.43031358885017423,
      "grad_norm": 0.5076520442962646,
      "kl": 0.04052734375,
      "learning_rate": 1e-06,
      "loss": 0.0016,
      "reward": 1.2500000596046448,
      "reward_std": 0.12974976375699043,
      "rewards/accuracy_reward": 0.2857142984867096,
      "rewards/format_reward": 0.9642857313156128,
      "step": 247
    },
    {
      "completion_length": 96.41071701049805,
      "epoch": 0.43205574912891986,
      "grad_norm": 2.3108022212982178,
      "kl": 0.0411376953125,
      "learning_rate": 1e-06,
      "loss": 0.0016,
      "reward": 1.5000001192092896,
      "reward_std": 0.11266788095235825,
      "rewards/accuracy_reward": 0.5000000149011612,
      "rewards/format_reward": 1.0,
      "step": 248
    },
    {
      "completion_length": 106.32143020629883,
      "epoch": 0.4337979094076655,
      "grad_norm": 2.105095386505127,
      "kl": 0.03216552734375,
      "learning_rate": 1e-06,
      "loss": 0.0013,
      "reward": 1.3392857313156128,
      "reward_std": 0.14838216453790665,
      "rewards/accuracy_reward": 0.3392857313156128,
      "rewards/format_reward": 1.0,
      "step": 249
    },
    {
      "completion_length": 74.57143020629883,
      "epoch": 0.4355400696864111,
      "grad_norm": 2.9153895378112793,
      "kl": 0.04345703125,
      "learning_rate": 1e-06,
      "loss": 0.0017,
      "reward": 1.4642857909202576,
      "reward_std": 0.1539071872830391,
      "rewards/accuracy_reward": 0.4642857313156128,
      "rewards/format_reward": 1.0,
      "step": 250
    },
    {
      "completion_length": 199.92858123779297,
      "epoch": 0.4372822299651568,
      "grad_norm": 2.763808488845825,
      "kl": 0.0286865234375,
      "learning_rate": 1e-06,
      "loss": 0.0011,
      "reward": 1.2321429252624512,
      "reward_std": 0.3324786275625229,
      "rewards/accuracy_reward": 0.2857142873108387,
      "rewards/format_reward": 0.9464285969734192,
      "step": 251
    },
    {
      "completion_length": 68.98214721679688,
      "epoch": 0.43902439024390244,
      "grad_norm": 1.7675226926803589,
      "kl": 0.0506591796875,
      "learning_rate": 1e-06,
      "loss": 0.002,
      "reward": 1.5178571939468384,
      "reward_std": 0.1071428656578064,
      "rewards/accuracy_reward": 0.517857164144516,
      "rewards/format_reward": 1.0,
      "step": 252
    },
    {
      "completion_length": 66.26785945892334,
      "epoch": 0.44076655052264807,
      "grad_norm": 2.443754196166992,
      "kl": 0.03668212890625,
      "learning_rate": 1e-06,
      "loss": 0.0015,
      "reward": 1.4821429252624512,
      "reward_std": 0.29123930633068085,
      "rewards/accuracy_reward": 0.4821428954601288,
      "rewards/format_reward": 1.0,
      "step": 253
    },
    {
      "completion_length": 216.76786041259766,
      "epoch": 0.4425087108013937,
      "grad_norm": 0.465847373008728,
      "kl": 0.03338623046875,
      "learning_rate": 1e-06,
      "loss": 0.0013,
      "reward": 1.3928572535514832,
      "reward_std": 0.20117833465337753,
      "rewards/accuracy_reward": 0.4464285969734192,
      "rewards/format_reward": 0.9464285969734192,
      "step": 254
    },
    {
      "completion_length": 107.25000381469727,
      "epoch": 0.4442508710801394,
      "grad_norm": 1.1041370630264282,
      "kl": 0.0469970703125,
      "learning_rate": 1e-06,
      "loss": 0.0019,
      "reward": 1.4821429252624512,
      "reward_std": 0.1896214708685875,
      "rewards/accuracy_reward": 0.5000000298023224,
      "rewards/format_reward": 0.9821428656578064,
      "step": 255
    },
    {
      "completion_length": 134.30358123779297,
      "epoch": 0.445993031358885,
      "grad_norm": 1.6369932889938354,
      "kl": 0.0478515625,
      "learning_rate": 1e-06,
      "loss": 0.0019,
      "reward": 1.3571429252624512,
      "reward_std": 0.1539071798324585,
      "rewards/accuracy_reward": 0.3750000149011612,
      "rewards/format_reward": 0.9821428656578064,
      "step": 256
    },
    {
      "completion_length": 127.5535774230957,
      "epoch": 0.44773519163763065,
      "grad_norm": 0.9092658162117004,
      "kl": 0.0428466796875,
      "learning_rate": 1e-06,
      "loss": 0.0017,
      "reward": 1.2678571939468384,
      "reward_std": 0.0357142873108387,
      "rewards/accuracy_reward": 0.267857164144516,
      "rewards/format_reward": 1.0,
      "step": 257
    },
    {
      "completion_length": 132.4107208251953,
      "epoch": 0.44947735191637633,
      "grad_norm": 1.8821191787719727,
      "kl": 0.042724609375,
      "learning_rate": 1e-06,
      "loss": 0.0017,
      "reward": 1.3750000596046448,
      "reward_std": 0.1785714402794838,
      "rewards/accuracy_reward": 0.392857164144516,
      "rewards/format_reward": 0.9821428656578064,
      "step": 258
    },
    {
      "completion_length": 197.5714340209961,
      "epoch": 0.45121951219512196,
      "grad_norm": 2.47094464302063,
      "kl": 0.02911376953125,
      "learning_rate": 1e-06,
      "loss": 0.0012,
      "reward": 1.2142857909202576,
      "reward_std": 0.2253357619047165,
      "rewards/accuracy_reward": 0.2321428693830967,
      "rewards/format_reward": 0.9821428656578064,
      "step": 259
    },
    {
      "completion_length": 134.96429443359375,
      "epoch": 0.4529616724738676,
      "grad_norm": 1.3298813104629517,
      "kl": 0.0323486328125,
      "learning_rate": 1e-06,
      "loss": 0.0013,
      "reward": 1.3750000596046448,
      "reward_std": 0.1181928962469101,
      "rewards/accuracy_reward": 0.3750000298023224,
      "rewards/format_reward": 1.0,
      "step": 260
    },
    {
      "completion_length": 104.3035774230957,
      "epoch": 0.4547038327526132,
      "grad_norm": 2.41137957572937,
      "kl": 0.044921875,
      "learning_rate": 1e-06,
      "loss": 0.0018,
      "reward": 1.071428656578064,
      "reward_std": 0.11266787722706795,
      "rewards/accuracy_reward": 0.07142857648432255,
      "rewards/format_reward": 1.0,
      "step": 261
    },
    {
      "completion_length": 111.57143783569336,
      "epoch": 0.4564459930313589,
      "grad_norm": 2.6853108406066895,
      "kl": 0.022705078125,
      "learning_rate": 1e-06,
      "loss": 0.0009,
      "reward": 1.3571429252624512,
      "reward_std": 0.1428571492433548,
      "rewards/accuracy_reward": 0.3571428805589676,
      "rewards/format_reward": 1.0,
      "step": 262
    },
    {
      "completion_length": 93.35714721679688,
      "epoch": 0.45818815331010454,
      "grad_norm": 2.3133461475372314,
      "kl": 0.0418701171875,
      "learning_rate": 1e-06,
      "loss": 0.0017,
      "reward": 1.5178571939468384,
      "reward_std": 0.2610500380396843,
      "rewards/accuracy_reward": 0.5357142984867096,
      "rewards/format_reward": 0.9821428656578064,
      "step": 263
    },
    {
      "completion_length": 168.1607208251953,
      "epoch": 0.45993031358885017,
      "grad_norm": 2.1456406116485596,
      "kl": 0.03265380859375,
      "learning_rate": 1e-06,
      "loss": 0.0013,
      "reward": 1.285714328289032,
      "reward_std": 0.2253357619047165,
      "rewards/accuracy_reward": 0.2857143059372902,
      "rewards/format_reward": 1.0,
      "step": 264
    },
    {
      "completion_length": 117.03571701049805,
      "epoch": 0.4616724738675958,
      "grad_norm": 2.640500545501709,
      "kl": 0.03973388671875,
      "learning_rate": 1e-06,
      "loss": 0.0016,
      "reward": 1.410714328289032,
      "reward_std": 0.14838216453790665,
      "rewards/accuracy_reward": 0.4107142984867096,
      "rewards/format_reward": 1.0,
      "step": 265
    },
    {
      "completion_length": 206.94644165039062,
      "epoch": 0.4634146341463415,
      "grad_norm": 0.7177957892417908,
      "kl": 0.02777099609375,
      "learning_rate": 1e-06,
      "loss": 0.0011,
      "reward": 1.1250000596046448,
      "reward_std": 0.1071428656578064,
      "rewards/accuracy_reward": 0.1428571492433548,
      "rewards/format_reward": 0.9821428656578064,
      "step": 266
    },
    {
      "completion_length": 181.1607208251953,
      "epoch": 0.4651567944250871,
      "grad_norm": 1.6306074857711792,
      "kl": 0.03717041015625,
      "learning_rate": 1e-06,
      "loss": 0.0015,
      "reward": 1.160714328289032,
      "reward_std": 0.29123931378126144,
      "rewards/accuracy_reward": 0.1785714402794838,
      "rewards/format_reward": 0.9821428656578064,
      "step": 267
    },
    {
      "completion_length": 165.6428680419922,
      "epoch": 0.46689895470383275,
      "grad_norm": 1.0887411832809448,
      "kl": 0.02734375,
      "learning_rate": 1e-06,
      "loss": 0.0011,
      "reward": 1.5357143878936768,
      "reward_std": 0.17553050816059113,
      "rewards/accuracy_reward": 0.5535714477300644,
      "rewards/format_reward": 0.9821428656578064,
      "step": 268
    },
    {
      "completion_length": 172.42858123779297,
      "epoch": 0.4686411149825784,
      "grad_norm": 0.8613088130950928,
      "kl": 0.03253173828125,
      "learning_rate": 1e-06,
      "loss": 0.0013,
      "reward": 1.160714328289032,
      "reward_std": 0.07695359364151955,
      "rewards/accuracy_reward": 0.1785714328289032,
      "rewards/format_reward": 0.9821428656578064,
      "step": 269
    },
    {
      "completion_length": 235.9821548461914,
      "epoch": 0.47038327526132406,
      "grad_norm": 1.2009068727493286,
      "kl": 0.0572509765625,
      "learning_rate": 1e-06,
      "loss": 0.0023,
      "reward": 1.1250000596046448,
      "reward_std": 0.21981072798371315,
      "rewards/accuracy_reward": 0.14285714365541935,
      "rewards/format_reward": 0.9821428656578064,
      "step": 270
    },
    {
      "completion_length": 155.21428680419922,
      "epoch": 0.4721254355400697,
      "grad_norm": 0.7742925882339478,
      "kl": 0.0279541015625,
      "learning_rate": 1e-06,
      "loss": 0.0011,
      "reward": 1.3214285969734192,
      "reward_std": 0.0714285746216774,
      "rewards/accuracy_reward": 0.321428582072258,
      "rewards/format_reward": 1.0,
      "step": 271
    },
    {
      "completion_length": 179.2678680419922,
      "epoch": 0.4738675958188153,
      "grad_norm": 1.1722646951675415,
      "kl": 0.02838134765625,
      "learning_rate": 1e-06,
      "loss": 0.0011,
      "reward": 1.4285715222358704,
      "reward_std": 0.0714285746216774,
      "rewards/accuracy_reward": 0.4285714477300644,
      "rewards/format_reward": 1.0,
      "step": 272
    },
    {
      "completion_length": 201.83929443359375,
      "epoch": 0.47560975609756095,
      "grad_norm": 1.630911946296692,
      "kl": 0.0638427734375,
      "learning_rate": 1e-06,
      "loss": 0.0025,
      "reward": 1.3035714626312256,
      "reward_std": 0.07695359364151955,
      "rewards/accuracy_reward": 0.321428582072258,
      "rewards/format_reward": 0.9821428656578064,
      "step": 273
    },
    {
      "completion_length": 126.08929061889648,
      "epoch": 0.47735191637630664,
      "grad_norm": 2.8433239459991455,
      "kl": 0.032958984375,
      "learning_rate": 1e-06,
      "loss": 0.0013,
      "reward": 1.3928572535514832,
      "reward_std": 0.18409645557403564,
      "rewards/accuracy_reward": 0.3928571492433548,
      "rewards/format_reward": 1.0,
      "step": 274
    },
    {
      "completion_length": 146.4464340209961,
      "epoch": 0.47909407665505227,
      "grad_norm": 2.8697118759155273,
      "kl": 0.02886962890625,
      "learning_rate": 1e-06,
      "loss": 0.0012,
      "reward": 1.4285714626312256,
      "reward_std": 0.11266787722706795,
      "rewards/accuracy_reward": 0.4285714328289032,
      "rewards/format_reward": 1.0,
      "step": 275
    },
    {
      "completion_length": 235.2321548461914,
      "epoch": 0.4808362369337979,
      "grad_norm": 1.3921054601669312,
      "kl": 0.0537109375,
      "learning_rate": 1e-06,
      "loss": 0.0022,
      "reward": 1.2142857909202576,
      "reward_std": 0.1428571529686451,
      "rewards/accuracy_reward": 0.2678571492433548,
      "rewards/format_reward": 0.9464285969734192,
      "step": 276
    },
    {
      "completion_length": 173.71429443359375,
      "epoch": 0.48257839721254353,
      "grad_norm": 1.9831897020339966,
      "kl": 0.093994140625,
      "learning_rate": 1e-06,
      "loss": 0.0037,
      "reward": 1.3214285969734192,
      "reward_std": 0.0714285746216774,
      "rewards/accuracy_reward": 0.3214285969734192,
      "rewards/format_reward": 1.0,
      "step": 277
    },
    {
      "completion_length": 191.69644165039062,
      "epoch": 0.4843205574912892,
      "grad_norm": 1.5223335027694702,
      "kl": 0.03350830078125,
      "learning_rate": 1e-06,
      "loss": 0.0013,
      "reward": 1.3035714626312256,
      "reward_std": 0.21981073915958405,
      "rewards/accuracy_reward": 0.321428582072258,
      "rewards/format_reward": 0.9821428656578064,
      "step": 278
    },
    {
      "completion_length": 170.60714721679688,
      "epoch": 0.48606271777003485,
      "grad_norm": 0.9542393088340759,
      "kl": 0.03057861328125,
      "learning_rate": 1e-06,
      "loss": 0.0012,
      "reward": 1.446428656578064,
      "reward_std": 0.1071428619325161,
      "rewards/accuracy_reward": 0.4642857313156128,
      "rewards/format_reward": 0.9821428656578064,
      "step": 279
    },
    {
      "completion_length": 236.96428680419922,
      "epoch": 0.4878048780487805,
      "grad_norm": 1.6516475677490234,
      "kl": 0.0150146484375,
      "learning_rate": 1e-06,
      "loss": 0.0006,
      "reward": 1.3392857313156128,
      "reward_std": 0.07695358991622925,
      "rewards/accuracy_reward": 0.3392857313156128,
      "rewards/format_reward": 1.0,
      "step": 280
    },
    {
      "completion_length": 201.25000762939453,
      "epoch": 0.4895470383275261,
      "grad_norm": 1.9830681085586548,
      "kl": 0.02423095703125,
      "learning_rate": 1e-06,
      "loss": 0.001,
      "reward": 1.3392857909202576,
      "reward_std": 0.1785714402794838,
      "rewards/accuracy_reward": 0.3571428805589676,
      "rewards/format_reward": 0.9821428656578064,
      "step": 281
    },
    {
      "completion_length": 186.05358123779297,
      "epoch": 0.4912891986062718,
      "grad_norm": 3.112156629562378,
      "kl": 0.02825927734375,
      "learning_rate": 1e-06,
      "loss": 0.0011,
      "reward": 1.3571429252624512,
      "reward_std": 0.26657506078481674,
      "rewards/accuracy_reward": 0.3571428805589676,
      "rewards/format_reward": 1.0,
      "step": 282
    },
    {
      "completion_length": 149.62500762939453,
      "epoch": 0.4930313588850174,
      "grad_norm": 0.9592561721801758,
      "kl": 0.03778076171875,
      "learning_rate": 1e-06,
      "loss": 0.0015,
      "reward": 1.3571429252624512,
      "reward_std": 0.0714285746216774,
      "rewards/accuracy_reward": 0.357142873108387,
      "rewards/format_reward": 1.0,
      "step": 283
    },
    {
      "completion_length": 213.50000762939453,
      "epoch": 0.49477351916376305,
      "grad_norm": 0.42550015449523926,
      "kl": 0.02685546875,
      "learning_rate": 1e-06,
      "loss": 0.0011,
      "reward": 1.321428656578064,
      "reward_std": 0.0714285746216774,
      "rewards/accuracy_reward": 0.357142873108387,
      "rewards/format_reward": 0.9642857313156128,
      "step": 284
    },
    {
      "completion_length": 197.5357208251953,
      "epoch": 0.4965156794425087,
      "grad_norm": 0.991173505783081,
      "kl": 0.0289306640625,
      "learning_rate": 1e-06,
      "loss": 0.0012,
      "reward": 1.3928571939468384,
      "reward_std": 0.0714285746216774,
      "rewards/accuracy_reward": 0.392857164144516,
      "rewards/format_reward": 1.0,
      "step": 285
    },
    {
      "completion_length": 133.76786422729492,
      "epoch": 0.49825783972125437,
      "grad_norm": 1.0027332305908203,
      "kl": 0.0306396484375,
      "learning_rate": 1e-06,
      "loss": 0.0012,
      "reward": 1.160714328289032,
      "reward_std": 0.1071428619325161,
      "rewards/accuracy_reward": 0.160714291036129,
      "rewards/format_reward": 1.0,
      "step": 286
    },
    {
      "completion_length": 90.62500762939453,
      "epoch": 0.5,
      "grad_norm": 2.177978277206421,
      "kl": 0.05548095703125,
      "learning_rate": 1e-06,
      "loss": 0.0022,
      "reward": 1.1607143878936768,
      "reward_std": 0.14838216826319695,
      "rewards/accuracy_reward": 0.16071429289877415,
      "rewards/format_reward": 1.0,
      "step": 287
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 287,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}